Activity
From 05/10/2012 to 06/08/2012
06/08/2012
- 11:38 PM Revision bfbed508 (ceph): librbd: use ImageCtx members instead of the old header in resize()
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 11:38 PM Revision 7376a773 (ceph): librbd: Update ImageCtx for new format
- Detect the format when an image is opened by the presence of the
original format header object. Use member variables ... - 11:38 PM Revision 4eb2138e (ceph): librbd: update ictx_refresh to work with both formats
- It now sets the member variables of ImageCtx so other functions
don't have to use the on-disk header. If the features... - 11:38 PM Revision 00f4e841 (ceph): librbd: check that the current snapid for a snap name matches
- Checking that it exists doesn't prevent you from having the snapshot
change out from under you in the following situa... - 11:38 PM Revision 8e51e060 (ceph): librbd: remove on-disk header argument from helper functions
- Make most of them take the parameters they actually use.
trim_image() now takes an ImageCtx, which means remove() mus... - 11:38 PM Revision f1d69639 (ceph): librbd: use cls_client functions for calling class methods
- Use the old or new methods make resize, snapshot add and snapsnhot
remove work with both old and new formats.
Signed... - 11:38 PM Revision ef38394f (ceph): librbd: make rename work with any header format
- Instead of interpreting the header, just copy all the data and
omap values from the original header to the newly name... - 11:38 PM Revision 127b4259 (ceph): librbd: rename md_oid parameters to header_oid
- This is more consistent with the rest of the code now,
and is a bit more clear.
Signed-off-by: Josh Durgin <josh.dur... - 11:38 PM Revision 5bc34f4a (ceph): librbd: validate order before creating an image
- The value must be passed, and it shouldn't be below 4k
(enforced by the command line tool already) or above the
range... - 11:38 PM Revision dcc7c963 (ceph): cls_rbd: add methods for interacting with the new header format
- Add the client side (cls_rbd_client) for testing as well.
librbd will use the functions in cls_rbd_client to interact... - 11:14 PM Bug #2524: librados crashed while connecting to cluster
- Thanks for the update. Yes, we do have different models, including a pool of set number of rados_t instances, etc. Bu...
- 10:37 PM Bug #2524: librados crashed while connecting to cluster
- Xiaopong Tran wrote:
> This is on my system:
> [...]
>
> Does it create a thread to every configured osd or only one... - 09:27 PM Bug #2524: librados crashed while connecting to cluster
- I bumped up the threads-max to:...
- 07:40 PM Bug #2524: librados crashed while connecting to cluster
- This is on my system:...
- 07:17 AM Bug #2524: librados crashed while connecting to cluster
- Sage Weil wrote:
> can you cat /proc/sys/kernel/threads-max ? on my system it's only 127837.
Yeah, for each libr... - 07:09 AM Bug #2524: librados crashed while connecting to cluster
- can you cat /proc/sys/kernel/threads-max ? on my system it's only 127837.
- 03:17 AM Bug #2524: librados crashed while connecting to cluster
- Ah, formatting... sorry...
- 03:15 AM Bug #2524: librados crashed while connecting to cluster
- Alright, more information. I was thinking, maybe it was the max number of open files, or the stack size is too low, s...
- 11:04 PM Feature #2496 (In Progress): reinstall pudgy
- 09:35 PM Revision 8c08482c (ceph): regression: fix new rados, rbd test yamls
- Don't start cluster twice!
- 09:31 PM Revision 60c684a4 (ceph): objclass: add indent settings header
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 09:31 PM Revision e0fda59d (ceph): objclass: pass strings as const references
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 09:31 PM Revision ba56d812 (ceph): cls_rgw: add indent settings header
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 09:31 PM Revision 5750e53e (ceph): test_librbd: don't include a .cc file
- Add it as a source like the rest of the gtests.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 09:31 PM Revision fe46a749 (ceph): class_debug: add indent settings header
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 09:31 PM Revision 715c1f13 (ceph): objclass: allow classes to log at different levels
- Add a logging level parameter to CLS_LOG, and a CLS_ERR that's
analagous to derr.
For now cls_rgw is all logging at ... - 09:31 PM Revision 23ebc093 (ceph): objclass: rename omap methods to be consistent with other interfaces
- The osd ops and the librados methods all distinguish between keys
and values. It's useful to be able to read a bunch ... - 09:31 PM Revision 49059b23 (ceph): objclass: add methods to get keys without values, and set multiple values
- These will be useful in the new rbd header format. This brings the
API into parity with the available OSD ops for oma... - 09:31 PM Revision 8c637f56 (ceph): vstart: debug osd classes when debugging is on
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 09:31 PM Revision 13aa578d (ceph): DBObjectMap: remove extra semicolon
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 09:31 PM Revision 70686c59 (ceph): librbd: remove useless ENOMEM checks
- There will be an exception if memory can't be allocated.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 09:31 PM Revision 92325d0f (ceph): cls_rbd: remove unused test_exec and snap_revert methods
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 09:31 PM Revision fefedc1b (ceph): cls_rbd: add indent settings header
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 09:03 PM Feature #2337 (Resolved): rgw and rados performance numbers
- 10:14 AM Feature #2337: rgw and rados performance numbers
- Actually, the specific sprint test is here:
https://docs.google.com/a/inktank.com/spreadsheet/ccc?key=0AnmmfpoQ1_9... - 09:53 AM Feature #2337: rgw and rados performance numbers
- Results are being posted here:
https://docs.google.com/a/inktank.com/folder/d/0B3mmfpoQ1_94amRLQW5YT3l3OG8/edit - 07:48 PM Revision b2793c42 (ceph): mon: require force argument for 'mds newfs ...'
- This is dangerous, make sure they know what they're doing. I bet <someone>
$1 that Greg asked me to do this original... - 06:55 PM Revision 6df344c7 (ceph): run rados, rbd api tests under thrashing
- 06:34 PM Revision 3eca3604 (ceph): Cleaned some lingering references to "verbose"
- Signed-off-by: caleb miles <caleb.miles@inktank.com>
- 06:18 PM Revision 3876dbfd (ceph): Merge branch 'wip-crush-tunables'
- Reviewed-by: Sam Just <sam.just@inktank.com>
- 06:17 PM Revision c0a02a4a (ceph): crushtool: refine tunables warning message
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:02 PM Revision 978d5419 (ceph): re-include assert after boost::pool
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:58 PM Revision c04de2b3 (ceph): crush: generate histogram of choose tries
- Optionally populate a histogram of choose descent attempts.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:58 PM Revision dba70eed (ceph): CrushTester: dump histogram of choose attempts
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:58 PM Revision eaad9845 (ceph): CrushTester: clean up output interface
- Multiple accessors. Init in ctor. Avoid temp vars in crushtool.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:58 PM Revision cfbe34d5 (ceph): crushtool: --show-* instead of --output-*
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:58 PM Revision 1a2959ed (ceph): crushtool: add --enable-unsafe-tunables option
- This is required to adjust tunables.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:58 PM Revision 740aa3df (ceph): crush: compile/decompile tunables
- Lines like
tunable <name> <value>
will adjust tunables. Right now only three are defined.
--enable-unsafe-tunabl... - 05:58 PM Revision c4336a3a (ceph): crushtool: arguments to adjust tunables
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:58 PM Revision 48737114 (ceph): CrushTester: optionally output bad mappings
- Optionally dump bad inputs to stdout.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:58 PM Revision 4abb53d4 (ceph): crush: fix leaf recursion if we already collided
- This just saves us some cycles, but does not effect placement results at
all.
Signed-off-by: Sage Weil <sage@inktank... - 04:42 PM rbd Bug #2534 (Resolved): librbd: make sure watch is established on same header version as initial re...
- Right now there's a race where it doesn't.
- 11:16 AM Bug #2533 (Duplicate): osd: watchers tracked by entity_name_t, not by cookie
- In the object info, watchers are tracked in a map<entity_name_t, watch_info_t>, but if there are multiple watchers fr...
- 10:43 AM Feature #1711 (Resolved): chef: multiple monitor support
- Works as of ceph-cookbook.git commit b5cc21bf5b9c3f59474a7dfe38e04ee01b584fa3 and ceph.git commit 7332e9c717fb627d51e...
- 10:12 AM rbd Feature #2531: rbd: fencing broken clients
- I talked to Sam about the combination of blacklisting, bad client writes, and changing primaries that we discussed an...
- 10:11 AM Linux kernel client Feature #26 (Rejected): statlite
- 10:09 AM Linux kernel client Cleanup #2093 (Resolved): ceph-client: messenger: the "to" parameter to read_partial() needs to go
- 10:08 AM Linux kernel client Bug #2395 (Resolved): kernel crash after unmap a rdb device while the cluster is down
- I'm going go assume this is running the older code and close it. If not, let us know!
- 10:06 AM rbd Bug #2478 (New): krbd: unmap on 3.4.0: scheduling while atomic...
- 10:04 AM Linux kernel client Feature #949 (Rejected): rbd: async writes, flush/barrier
- 10:04 AM Linux kernel client Bug #2243 (Resolved): btrfs: warning in orphan_commit_root
- 09:51 AM rbd Bug #2532: rbd command allows passing in -K </path/to/secret>, but long version of (--secret) doe...
- That's probably best. It is always easier though when all subcommands under the main command, rbd in this case used o...
- 09:00 AM rbd Bug #2532: rbd command allows passing in -K </path/to/secret>, but long version of (--secret) doe...
- Oh, i see.
I think the right fix is to make '--secret' and synonym for '--keyfile', and fix up rbd to use the conf... - 08:20 AM rbd Bug #2532: rbd command allows passing in -K </path/to/secret>, but long version of (--secret) doe...
- When I try to use --keyfile=<file> with map, it seemingly fails, but using --secret=<file> succeeds. ...
- 08:13 AM rbd Bug #2532: rbd command allows passing in -K </path/to/secret>, but long version of (--secret) doe...
- This is part of the rbd cmd helper message. It seems that for the map command one uses --secret....
- 07:00 AM rbd Bug #2532: rbd command allows passing in -K </path/to/secret>, but long version of (--secret) doe...
- the option is --keyfile <file>... where did you see --secret <file> documented?
- 05:49 AM rbd Bug #2532 (Resolved): rbd command allows passing in -K </path/to/secret>, but long version of (--...
- While rolling back a snapshot I succeed when I pass in `-K with location of key file, but it looks like I fail when I...
06/07/2012
- 10:57 PM Revision 8b796974 (ceph): crush: make magic numbers tunable
- We have three magic numbers in crush_choose that are now tunable. The
first two control the local retry behavior, in... - 09:38 PM rbd Feature #2531 (Resolved): rbd: fencing broken clients
- 08:57 PM Revision 7332e9c7 (ceph): mon: use mode 0600 throughout
- Fixes: #2526
Signed-off-by: Sage Weil <sage@inktank.com> - 08:50 PM Revision 4fa665c1 (ceph): --summary: add total counts, also note free machines
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com> - 07:22 PM Revision 80939977 (ceph): Merge remote-tracking branch 'gh/mon-auth'
- Reviewed-by: Greg Farnum <greg@inktank.com>
- 06:45 PM Bug #2524: librados crashed while connecting to cluster
- objdump on the NIF shared library.
- 06:29 PM Bug #2524: librados crashed while connecting to cluster
- This is weird, if the problem is caused by resource exhaustion. I run this app on a machine with i7 CPU (with 8 cores...
- 09:24 AM Bug #2524: librados crashed while connecting to cluster
- This assert means that either a malloc or a call to pthread_create failed. It's probably resource exhaustion of some ...
- 04:23 AM Bug #2524 (Won't Fix): librados crashed while connecting to cluster
- Librados crahsed while connecting to the cluster.
Here is some log information. Unfortunately, I don't have more i... - 06:35 PM Revision 07169d2e (ceph): doc: Added mount cephfs with fstab.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 06:21 PM Revision 7d1b32a0 (ceph): osd: include past_intervals in pg query results
- This will help us figure out *why* nodes are in the prior set.
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-... - 06:21 PM Revision 36a39795 (ceph): OSD: _have_pg should return NULL if pg is not in map
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 06:21 PM Revision ff470104 (ceph): Merge remote-tracking branch 'gh/wip-assert2'
- "So be it"
Reviewed-by: Sam Just <sam.just@dreamhost.com> - 05:19 PM Revision fae1d47a (ceph): deliberately break encoding macros when wrong assert is present
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:18 PM Revision 3571f43c (ceph): misc assert #include cleanup, hackery
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:18 PM Revision a1ae8b6a (ceph): reinclude assert.h after json_spirit
- json_spirit clobbers it!
Signed-off-by: Sage Weil <sage@inktank.com> - 05:08 PM Revision e87a66b7 (ceph): doc: Incorporated Sam's comments.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 04:41 PM Revision 36658552 (ceph): Merge remote-tracking branch 'gh/wip-assert'
- Reviewed-by: Sam Just <sam.just@inktank.com>
- 04:25 PM rbd Documentation #2530 (Closed): Doc: rbd manpage doesn't mention watch; usage does, and it works
- 04:20 PM Tasks #2529 (Resolved): debian: Merge packaging changes from Ubuntu 12.04
- The package in ubuntu is split to ceph-fs-common (mount helpers), ceph-mds (not in main), etc. Merge what makes sense.
- 03:10 PM rbd Bug #2528 (Resolved): Mounted RBD image appears to go read-only after a snapshot is created
- I have been able to repeat this a number of times. Essentially, I create a small rbd device, using the map command in...
- 02:38 PM Revision 3939839b (ceph): doc: Typo fix.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 01:54 PM Bug #2526 (Resolved): ceph-mon $mon_data_dir/keyring is world readable
- gah... commit:7332e9c717fb627d51efcaa3f31473a2c129e876
- 01:25 PM Bug #2526 (Resolved): ceph-mon $mon_data_dir/keyring is world readable
- Keys to the kingdom, for anyone to grab. ceph-mon --mkfs creates this file, it should enforce the access mode.
ubu... - 01:52 PM rgw Bug #2527 (Resolved): RGW may return 409 Conflict when deleting a bucket
- If a bucket delete call occurs immediately after running a delete operation on the final remaining object in that buc...
- 12:53 PM Bug #2525 (Resolved): librados: some functions are not thread-safe
- Some functions are accessing the osdmap without any locks. There are probably other cases like this. Find and fix all...
- 03:29 AM Revision 44374bc4 (ceph): new variable lock hid lock() function
- 02:19 AM Revision 22863c30 (ceph): mon: set policy for client, mds before throttler
- Otherwise we fail the assert in Messenger::set_policy_throttler()!
Signed-off-by: Sage Weil <sage@inktank.com> - 02:10 AM Revision 71c2877c (ceph): DBObjectMap: fix some warnings
- os/DBObjectMap.cc:197: warning: suggest a space before ';' or explicit braces around empty body in 'for' statement
S... - 01:54 AM Revision 9313cdea (ceph): teuthology-lock: add --summary and --brief options
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
06/06/2012
- 11:57 PM Revision 0f6d90c7 (ceph): make everyone use our assert #include and macro
- ...as detected by the previous patch.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:07 PM Revision a30601a1 (ceph): assert: detect when /usr/include/assert.h clobbers us
- The normal assert.h is very rude in that it clobbers any existing assert
define and replaces it with its own. An sad... - 11:01 PM Revision 2be716cf (ceph): Merge remote-tracking branch 'gh/wip-msgr-interface'
- Reviewed-by: Sage Weil <sage@inktank.com>
- 11:00 PM Revision 9ec28433 (ceph): pull s3-tests.git using git, not http
- 11:00 PM Revision 7d5c7ee8 (ceph): pull s3-tests.git using git, not http
- 10:30 PM Revision 165fbd13 (ceph): keyserver: also authenticate against mon keyring
- If we don't have a secret, also check in the extra_secrets keyring.
This means we can also authenticate as any users... - 10:26 PM Revision 5d520f17 (ceph): keyring: implement get_caps()
- Simple accessor, mirrors KeyServerData.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:26 PM Revision ca433f28 (ceph): mon: share mon keyring with KeyServer
- This will let us authenticate against items in the mon keyring, like the
mon. key itself.
Signed-off-by: Sage Weil <... - 09:11 PM Revision 35a61af1 (ceph): Merge remote-tracking branch 'gh/wip_hobject_wpool'
- Reviewed-by: Sage Weil <sage@inktank.com>
- 09:09 PM Revision 47b202ec (ceph): mon: put cluster log at /var/log/ceph/$cluster.log and/or send to syslog
- Also, stop breaking it down by event severity on disk. If you want that,
use syslog.
Fixes: #2497
Backport: dho
Sig... - 09:07 PM Feature #1422 (Resolved): libvirt: rbd storage pool
- 09:06 PM Feature #2486 (Resolved): crush: evaluate local retry behavior
- 09:06 PM Feature #2493 (Resolved): teuthology-lock --status
- 09:05 PM devops Feature #2498 (Fix Under Review): standardize keyring locations for daemons
- 08:32 PM Revision 7523ff3e (ceph): ceph: simplify 'cluster' mon log handling
- It's not a special file in the mon_data directory anymore, but intead
something in archive that will get slurped up n... - 06:29 PM Revision 37705edb (ceph): Merge branch 'wip-crushtool'
- Reviewed-by: Sage Weil <sage@inktank.com>
- 06:28 PM Revision 78b0bea1 (ceph): monclient: be paranoid/defensive about send_log vs log_client==NULL
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:07 PM Revision 58b02f93 (ceph): crushtool: fix cli tests given new less-chatty output, help
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:07 PM Revision af4d8db5 (ceph): crushtool: allow user to select output reporting in blocks
- Signed-off-by: caleb miles <caleb.miles@inktank.com>
- 05:45 PM Revision 87327c2d (ceph): doc: Added mount cephfs and included it in quick start.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 03:57 PM Messengers Cleanup #2150 (Resolved): repair the Simple/Messenger interface
- 02:06 PM Feature #2497 (Resolved): mon: new cluster logging strategy
- commit:47b202ecfdc00996b085a0c0d557564fbaa8bdfe
- 12:28 PM Feature #2497 (Fix Under Review): mon: new cluster logging strategy
- 12:28 PM Feature #2497: mon: new cluster logging strategy
- see wip-2497
- 01:27 PM Linux kernel client Bug #2523 (Resolved): xfs: xfs_iolock_reclaimable
- ...
- 01:22 PM rbd Bug #2522: xfstest #219
- ubuntu@teuthology:/a/nightly_coverage_2012-06-05-b
- 01:21 PM rbd Bug #2522 (Closed): xfstest #219
- ...
- 11:30 AM Bug #2518 (Resolved): mon: limit size of paxos log event
- 11:29 AM RADOS Feature #2521: crush: control bucket vs device mark-down probabilities independently
- 11:27 AM RADOS Feature #2521 (Resolved): crush: control bucket vs device mark-down probabilities independently
- --mark-down-ratio -- probability that a device (in eligible bucket) will be marked down
--mark-down-bucket... - 11:27 AM RADOS Feature #2421 (Resolved): crush: quantitatively validate mapping quality
- 09:16 AM Bug #2520 (Duplicate): iozone random read/write with 4k block size hangs
- http://thread.gmane.org/gmane.comp.file-systems.ceph.devel/6777/focus=6856
User reports iozone random read/write (... - 04:27 AM Revision 0167fd0b (ceph): mon: include all types of stuck pgs in health detail
- We were just including the last one, which isn't as helpful.
Backport: dho
Signed-off-by: Sage Weil <sage@inktank.com> - 04:27 AM Revision d4c30da7 (ceph): mon: include pg acting in health detail
- Backport: dho
Signed-off-by: Sage Weil <sage@inktank.com> - 04:27 AM Revision 6c1558ec (ceph): logclient: fix warning
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:20 AM Bug #2508: osdc/ObjectCacher.cc:761: void ObjectCacher::bh_write_commit(int64_t, sobject_t, loff_...
- Hi Josh,
i've increased osd_min_pg_log_entries to 5000. Let's see if it fixes the problem.
Simon - 03:16 AM Revision 37bf2a14 (ceph): test/cli/ceph-authtool: keyring.bin -> keyring
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:16 AM Revision 5b443ebd (ceph): doc: keyring.bin -> keyring everwhere
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:41 AM Revision 120ce3f8 (ceph): Pass up unmodified exceptions from connection.connect()
- This allows useful errors to be reported from things like
mismatched hostkeys, etc. - 01:40 AM Revision fac88a40 (ceph): More shortnames fixes:
- - Allow shortnames in teuthology-updatekeys as well
- Use list comprehensions instead of map() - 12:47 AM Revision f918049b (ceph): test/: Made omap_bench compatible with teuthology
- added --name parsing, made histogram better, made rados_id
configurable, changed object names to use configurable pre... - 12:09 AM Revision 02a9a013 (ceph): doc: Added the root discussion to deploy with mkcephfs.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 12:08 AM Revision 993caf81 (ceph): doc: Added chmod for keyring, and moved client.admin user higher.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
06/05/2012
- 11:39 PM Revision 85fee040 (ceph): logclient: fix crashes, fix which entries are sent
- I was seeing crashes when the monitor tried to send log entries.
* Send log entries that haven't already been sent.
... - 11:39 PM Revision b6059ccc (ceph): monclient: send more log entries when first set is acked
- Immediately send more log messages if we had more when the first set was
sent. Otherwise, wait until the next tick t... - 11:39 PM Revision 235f9274 (ceph): logclient: not a dispatcher
- Let MonClient and Monitor handle delivery of messages. This puts them in
control and lets them trigger sending of mo... - 11:38 PM Revision dfd9f5f5 (ceph): logclient: limit messages per MLog message
- This will avoid sending huge chunks of entries to the monitor and making
its life difficult.
Signed-off-by: Sage Wei... - 11:38 PM Revision 540865f4 (ceph): mon: limit size of each logm paxos event
- Limit the number of log events we cram into a single paxos event.
Fixes: #2518
Signed-off-by: Sage Weil <sage@inktan... - 11:33 PM Revision ee945128 (ceph): config_opts: filestore_update_to defaults to 1000
- This way, filestores will be auto-upgraded by default.
Signed-off-by: Samuel Just <sam.just@inktank.com> - 11:09 PM Revision 01ad679a (ceph): DBObjectMap: update check() for new format
- Signed-off-by: Samuel Just <sam.just@dreamhost.com>
- 11:09 PM Revision 19c667fb (ceph): DBObjectMap::init: initialize seq and v to correct values
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:09 PM Revision 6ec74acd (ceph): ReplicatedPG: remove_object_with_snap_hardlinks before creating temp obj
- hobject_ts must be unique in the filestore. Thus, when we create the
new temp object, the old one must have been del... - 11:09 PM Revision 7c8aac7b (ceph): ReplicatedPG: adjust missing at push_start
- When we start recieving an object, we remove the old copy. This will
prevent the primary from using that old copy af... - 11:09 PM Revision 92807c21 (ceph): OSD: exit(0) once filestore is converted
- Also, do not upgrade filestore automatically
Signed-off-by: Samuel Just <sam.just@dreamhost.com> - 11:09 PM Revision df7c4072 (ceph): FileStore: skip omap during list_collections
- Signed-off-by: Samuel Just <sam.just@dreamhost.com>
- 11:09 PM Revision ebfde20a (ceph): DBObjectMap: update header comments for new structure
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:09 PM Revision 875bec29 (ceph): DBObjectMap,FileStore: Remove IndexedPath parameters from ObjectMap
- IndexedPath parameters are no longer needed for getting the object
collections or for supporting the TMAP implementat... - 11:09 PM Revision 56cf461e (ceph): OSD,FileStore: clean up filestore convsersion
- Previously, we messed with the filestore_update_collections config
option to enable upgrades in the filestore. We no... - 11:09 PM Revision c64fc291 (ceph): test_object_map: remove DBObjectMapv0
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:09 PM Revision a2eedec5 (ceph): osd_types,PG: handle new hobject format in object_info,pg_log
- There are also legacy hobject encodings in the pg log and in object_info
attributes on objects.
Signed-off-by: Samue... - 11:09 PM Revision 5250e247 (ceph): test_filestore_idempotent_sequence: add omap
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:09 PM Revision ffc5f7ef (ceph): ReplicatedPG: push_start, don't insert empty extent into data_subset
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:09 PM Revision 50331ab1 (ceph): FileStore,DBObjectMap: add SequencerPosition argument to ObjectMap
- Previously, sequences like:
1. touch (c1, a)
2. link (c1, c2, a)
3. rm (c1, a)
4. setattr (c2, a)
5. clone (c2, a, b... - 11:09 PM Revision d8f8b1fa (ceph): OSD: do not convert an entire collection in one transaction
- Previously, we atomically moved the collection out of the way, created a
new collection, moved the contents of the ol... - 11:09 PM Revision d5ab8779 (ceph): src/: Add namespace and pool fields to hobject_t
- From this point, hobjects in the ObjectStore will be globally unique. This
will allow us to avoid including the coll... - 11:09 PM Revision 927458cf (ceph): test/ObjectMap: Copy current DBObjectMap implementation
- This implementation will be used to test the upgrade process.
Signed-off-by: Samuel Just <sam.just@dreamhost.com> - 11:09 PM Revision dab238b0 (ceph): os/: update CollectionIndex filename encodings
- filename encodings now include namespace and pool.
Signed-off-by: Samuel Just <sam.just@inktank.com> - 11:09 PM Revision 7a3a2d24 (ceph): DBObjectMap: add parse method for old encoding
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:09 PM Revision 9ceed879 (ceph): DBObjectMap: version bump for new format
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:09 PM Revision 761140a0 (ceph): FileStore,DBObjectMap: remove ObjectMap link method
- hobject_t's are now globally unique in filestore. Essentially, there is
a 1-to-1 mapping from inodes to hobject_t's.... - 11:09 PM Revision e083e0e9 (ceph): DBObjectMap: restructure for unique hobject_t's
- Previously, the ObjectStore operated in terms of (coll_t,hobject_t)
tupples. Now that hobject_t's are globally uniqu... - 11:09 PM Revision b6a99321 (ceph): DBObjectMap: Implement upgrade from previous format
- Also includes tests in test_object_map.cc
Signed-off-by: Samuel Just <sam.just@dreamhost.com> - 10:30 PM Revision 23c72930 (ceph): task/: Added object map benchmarking test
- Signed-off-by: Eleanor Cawthon <eleanor.cawthon@inktank.com>
- 06:57 PM Revision ec689e3e (ceph): rgw: replace 'should_log' with 'enable_ops_log'
- code cleanup, make it more clear about log intention.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 06:55 PM Revision 8e41ac15 (ceph): mon: clear osd_stat on osd creation/destruction
- Reported-by: Vladimir Bashkirtsev <vladimir@bashkirtsev.com>
Signed-off-by: Sage Weil <sage@inktank.com> - 06:09 PM Revision fb6ea82f (ceph): doc: Added S3 examples to the toctree.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 05:55 PM Revision 62fe9f42 (ceph): doc: adding code samples for S3 API usage (thanks, DH!)
- Signed-off-by: Ross Turk <ross@inktank.com>
- 05:40 PM Revision 644a615d (ceph): Makefile.am: explicitly mention that -Wl,--as-needed is location-sensit...
- 03:15 PM Revision e83a84c0 (ceph): doc: Added ${lsb_release -sc} based on Sam's feedback.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 01:36 PM Feature #2519 (Resolved): rados: allow setting pg_num and pgp_num when creating a pool
- Right now rados mkpool creates a pool with 8 pgs, which is almost always too few. 'ceph osd pool create' accepts pg_n...
- 01:04 PM Bug #2518: mon: limit size of paxos log event
- 01:03 PM Bug #2518 (Resolved): mon: limit size of paxos log event
- dho was having trouble with a 400MB paxos event/record. make LogMonitor limit an individual paxos event to something...
- 11:42 AM rgw Feature #2517 (Resolved): rgw: limit number of buckets per user (configurable per user)
- 11:37 AM rgw Feature #2516 (Resolved): rgw: new bandwidth-only per-user log
- - orthogonal to operations logs
- only aggregate user bandwidth usage (read, write) per date
- rgw sends a perio... - 11:02 AM Bug #2508: osdc/ObjectCacher.cc:761: void ObjectCacher::bh_write_commit(int64_t, sobject_t, loff_...
- Hi Simon,
If this is at all reproducible, could you try setting osd_min_pg_log_entries higher on all your osds, sa... - 07:47 AM Bug #2508 (Resolved): osdc/ObjectCacher.cc:761: void ObjectCacher::bh_write_commit(int64_t, sobje...
- Hi,
we've random KVM VPS crashes with the following error:... - 10:32 AM Feature #2510: update on-disk hobject_t encoding to include pool and namespace fields
- 10:15 AM Feature #2510 (Resolved): update on-disk hobject_t encoding to include pool and namespace fields
- This will allow hobject_t's to be globally unique in the filestore. That is, there will be a 1-to-1 inode to hobject...
- 10:31 AM Subtask #2515: allow collection upgrade to use more than one transaction
- 10:31 AM Subtask #2515 (Resolved): allow collection upgrade to use more than one transaction
- 10:31 AM Subtask #2514: Implement DBObjectMap upgrade from old version
- 10:30 AM Subtask #2514 (Resolved): Implement DBObjectMap upgrade from old version
- 10:31 AM Subtask #2513: Update DBObjectMap implementation to ignore collection
- 10:30 AM Subtask #2513 (Resolved): Update DBObjectMap implementation to ignore collection
- This allows us to remove the (coll_t,hobject_t)->seq mapping and directly store the leaf nodes keyed by hobject_t.
- 10:31 AM Subtask #2512: implement upgrade process for collections
- 10:29 AM Subtask #2512 (Resolved): implement upgrade process for collections
- also upgrade object_info and pg log encodings
- 10:31 AM Subtask #2511: Change hobject_t encoding
- 10:16 AM Subtask #2511 (Resolved): Change hobject_t encoding
- 10:17 AM CephFS Bug #733: cmds crash: mds/LogEvent.cc:88: FAILED assert(p.end())
- ok here is a logfile with the following config:
[mds]
debug = 20
debug ms = 1
debug md... - 10:08 AM Subtask #2402 (In Progress): audit calls into osd from pg for locking correctness
- 10:07 AM Subtask #2509 (Resolved): create OSDService to limit pg/osd interface
- 10:06 AM Subtask #2430: simplify pg removal
- 10:06 AM Subtask #2403: remove osd pointer from PG
- 10:06 AM Subtask #2333: create queueing for peering messages
- 10:06 AM Subtask #825: osd: remove pg map updating from handle_osd_map
- 10:06 AM Subtask #2332: move pg queueing into pgs
- 10:06 AM Subtask #2282: Handle map updates on a per-pg basis
- 09:56 AM rbd Feature #1480: librbd: image locking
- lock(entity)
unlock(entity)
new code should lock before open, unlock on close.
the rbd map tool have 'lock lis... - 12:50 AM Revision 044697d1 (ceph): Allow short names to teuthology-lock (e.g. "plana14")
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Sam Just <sam.just@inktank.com> - 12:28 AM CephFS Bug #1047: mds: crash on anchor table query
- No, I am not sure about that. Only saw the same assert message and a similar trace, so I assumed it to be the same bug.
06/04/2012
- 11:15 PM Revision fc1f9e0b (ceph): rgw: shutdown init_timer
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 11:12 PM Revision 5087997a (ceph): rgw: try to create fcgi socket through open() first
- FCGX_OpenSocket might just exit() without any warning if it fails
to create the socket.
Signed-off-by: Yehuda Sadeh ... - 04:09 PM rgw Bug #2503 (Resolved): rgw: ungraceful failure when cannot create unix domain socket
- Fixed, commit:5087997a1c90ecd1244dc1047a17858607c940f9.
- 03:09 PM rgw Bug #2503: rgw: ungraceful failure when cannot create unix domain socket
- No, another problem. This refers to the 'rgw socket path' that is being used for fastcgi.
- 06:26 AM rgw Bug #2503: rgw: ungraceful failure when cannot create unix domain socket
- There was a stupid error in master for a few days that was making noise about the admin socket.. is that what this wa...
- 03:56 PM Bug #2507 (Resolved): auth: "ceph auth get-or-create-key" argument validation is lacking
- This should probably have errored out:
ubuntu@inst01:~$ sudo ceph auth get-or-create-key client.foo borkbork
AQBW... - 01:08 PM CephFS Bug #1047: mds: crash on anchor table query
- Amon, are you sure you're hitting exactly this bug with your users? This particular one requires hard links to be in ...
- 01:04 PM CephFS Bug #733: cmds crash: mds/LogEvent.cc:88: FAILED assert(p.end())
- Aww, the actual debug line that's interesting here is generic_dout().
Can you do it again, this time adding "debug =... - 10:05 AM Messengers Cleanup #2150: repair the Simple/Messenger interface
- I scheduled another test run but I don't anticipate any problems — this should be reviewed for merge!
- 09:23 AM CephFS Bug #2494: mds: Cannot remove directory despite it being empty.
- Note that this was triggered frequently by backuppc runs:
http://thread.gmane.org/gmane.comp.file-systems.ceph.devel... - 09:23 AM Linux kernel client Bug #2506: ceph: ceph_add_cap: couldn't find snap realm NNN
- Note that this was triggered frequently by backuppc runs:
http://thread.gmane.org/gmane.comp.file-systems.ceph.devel... - 06:33 AM Bug #2487 (Resolved): rgw: (re)creating a suspended bucket succeeds
- 06:29 AM Bug #2491 (Resolved): watch/notify: racing notify and unwatch
- 04:31 AM Revision 000f9d30 (ceph): msg: make clear_pipe work only on a given Pipe, rather than the current...
- This way old Pipes that have been replaced can't clear the new Pipe
out of a Connection's link.
We might attempt to i... - 01:35 AM Bug #2346: xfs filesystem on top of rbd volume corrupts
- I am not 100% sure but it looks like kernel 3.2.17-1 fixed the problem. Let's wait 4 weeks to make sure of it.
06/03/2012
- 11:12 PM Revision dde85a20 (ceph): CrushTester: allow build without boost stuff for chi^2 testing
- With limited functionality.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:54 PM Revision 11b24f7c (ceph): crush: simiulate using rng; use chi-squared to measure result
- Signed-off-by: Caleb Miles <caleb.miles@inktank.com>
- 10:25 PM Revision aeb3a8f4 (ceph): crush: check_item_present
- True if id is present in the map.
Signed-off-by: Caleb Miles <caleb.miles@inktank.com> - 10:21 PM Revision 4061ae23 (ceph): ceph_argparse: with_float
- Signed-off-by: Caleb Miles <caleb.miles@inktank.com>
- 08:51 PM Revision a1c5ef04 (ceph): admin_socket: only init if path is defined
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:41 PM Linux kernel client Bug #2506 (Resolved): ceph: ceph_add_cap: couldn't find snap realm NNN
- From Martin Wilderoth on ML:...
- 05:39 PM CephFS Bug #2494: mds: Cannot remove directory despite it being empty.
06/02/2012
- 10:19 PM Revision 3764ca61 (ceph): upstart: simplify start; allow group stop via an abstract job
- Use a 'ceph-mds' or 'ceph-mon' event to start instances instead of
explicitly calling start. This avoids the ugly is... - 09:49 PM Revision ac28c260 (ceph): Merge remote branch 'gh/chef-3'
- 09:29 PM Revision c48de3ad (ceph): paxos: warn on extreme clock skew
- This would have helped us diagnose #2480.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:03 PM Revision 576dff89 (ceph): omapbench: fix warning
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:18 PM Revision ab5139b3 (ceph): omapbench: fix misc warnings
- Signed-off-by: Sage Weil <sage@inktank.com>
- 07:10 PM Feature #2003 (Rejected): limit XFS extent fragmentation for rbd
- 07:09 PM Cleanup #2435 (Resolved): Remove binary keyring support
- 07:09 PM Feature #2450 (Resolved): dho git commit emails
- 06:35 PM Feature #2505 (Resolved): Automatically set filestore xattr use omap
- It would be really nice if the filestore could automatically figure out if it should set "xattr use omap" depending o...
- 03:45 PM Feature #2414 (Resolved): upstart: support mds
- updated, tested, merged
- 03:26 PM Bug #2480 (Rejected): mon: election won't complete
- 03:20 PM Bug #2480: mon: election won't complete
- this was just a skewed clock. making the log warnings louder.
06/01/2012
- 11:13 PM Revision 6f3e7a84 (ceph): ReplicatedPG: fix pgls listing, add max listing size
- Previously, a client requesting a large pgls could tie up the
osd for an unacceptable amount of time. Also, it's pos... - 11:11 PM Revision d8e07a34 (ceph): objecter: fix pgls
- First problem: if the osd returns more entries than we ask for, max_entries
was going negative, and we were requestin... - 08:57 PM Revision 6eb99d48 (ceph): Merge remote-tracking branch 'gh/wip-admin'
- 08:54 PM Revision b9ef3149 (ceph): mon: fix slurp latest race
- It is possible for the latest version to get out in front of the
last_committed version:
a- start slurping
a- slur... - 08:46 PM Revision f871d837 (ceph): Makefile: include ceph-mds upstart bits in dist tarball
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:02 PM Revision 7a4e9234 (ceph): test/: Added object map benchmarking tool
- omap_bench writes configurable objectmaps to a configurable number
of objects and generates latency statistics.
Sign... - 07:53 PM Revision e0f19525 (ceph): doc: fix autobuild debian source line
- Signed-off-by: Sage Weil <sage@inktank.com>
- 07:43 PM rgw Bug #2504 (Resolved): rgw: use multiple notifications objects
- In order to spread the load.
- 07:35 PM rgw Bug #2503 (Resolved): rgw: ungraceful failure when cannot create unix domain socket
- At least need to add a log message before crashing.
- 04:44 PM Revision 97d43964 (ceph): mon: throttle client msgr memory
- Limit the amount of memory that can be consumed by client messages, similar
to the OSD. Do not limit inter-mon messa... - 04:42 PM Bug #2502 (Resolved): osd: do not mark self back up if internal heartbeats are failing
- saw this on congress. blocked up peering on teh cluster bc the osd was up but not doing anything useful due to clogg...
- 04:38 PM Bug #2501 (Won't Fix): acting.size() < nreplicas shouldn't prevent trimming old maps
- Currently, we won't mark a pg clean if we have too few replicas. This is correct, but it means we won't trim maps wh...
- 04:38 PM Bug #2500 (Resolved): osd: unprotected ::decodes in ReplicatedPG::do_osd_ops
- Decoding of data supplied by a client should not be able to crash the OSD. These should each be in a try/catch block.
- 04:30 PM Revision 52b28ed6 (ceph): Merge remote-tracking branch 'origin/wip-2491'
- 03:35 PM rgw Feature #2499 (Resolved): rgw: ability to delete users without first emptying and deleting all bu...
- Currently there is no way to simply delete a user without first going through and emptying then deleting its buckets....
- 01:51 PM Bug #2379 (Resolved): Mon crash after start
- commit:b9ef3149f8639e1f3cdbbb81eecb05e42be00711
- 10:42 AM devops Feature #2498 (Resolved): standardize keyring locations for daemons
- Currently keyring is controlled by g_conf->keyring, with a default search path that makes sense for clients but not m...
- 10:40 AM Feature #2497 (Resolved): mon: new cluster logging strategy
- * putting the log files in $mon_data is probably a bad idea
* they don't get rotated
* they aren't in a sufficientl... - 10:39 AM Feature #2496 (Resolved): reinstall pudgy
- 09:41 AM Feature #2495 (Resolved): mon: throttle incoming non-monitor msgr traffic
- commit:97d4396415a95f62f6deccda62d105d8d04717a4
- 09:02 AM Feature #2495 (Resolved): mon: throttle incoming non-monitor msgr traffic
- 04:39 AM Revision d3f855ec (ceph): fix up dist var
- This lets you override the default (now precise) in the ceph config yaml,
e.g.
- ceph:
dist: oneiric
branch:... - 04:21 AM Bug #2267: Ceph client crashed after shutting down one mds and osd
- Alex Elder wrote:
> It looks to me that in order for msg->bio_iter to be null, the
> value of msg->bio itself must ... - 02:05 AM Revision 7e3d90ac (ceph): Objecter: tone down linger op messages on tick
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 12:09 AM Revision af4fe154 (ceph): Change hardcoded oneiric to precise
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 12:04 AM Revision 4a7683cb (ceph): test_stress_watch: exercise watch/unwatch/close from second client
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
05/31/2012
- 11:46 PM Revision 95e0a880 (ceph): qa: stress_watch.sh workunit runs test_stress_watch
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:44 PM Revision 95ecf40e (ceph): add rados_stress_watch to regression
- 11:42 PM Revision a4c90b72 (ceph): Merge remote-tracking branch 'gh/wip-mon-doc'
- 11:01 PM Bug #2379 (Fix Under Review): Mon crash after start
- see wip-2379
- 10:46 PM Bug #2379: Mon crash after start
- triggered this on congress.
looks like latest is a 150 epochs out ahead of last_committed and the actual states:
<pr... - 10:35 PM Revision fb7ce59b (ceph): doc: Added preliminary rbd and ko info.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 09:20 PM Revision 10b0db3b (ceph): osd: Adding const to methods in OSDMap
- Signed-off-by: Adam Crume <adamcrume@gmail.com>
Signed-off-by: Sage Weil <sage@inktank.com> - 06:38 PM Revision 8609caff (ceph): vstart.sh: initialize keyring even if not using cephx
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 06:37 PM Revision e929d54d (ceph): Merge branch 'master' of ssh://github.com/ceph/ceph
- 02:40 PM CephFS Bug #2494 (Resolved): mds: Cannot remove directory despite it being empty.
- Running ceph 0.47.1 on top of XFS I've got at least two directories which used to contain files but are now empty and...
- 01:47 PM Feature #2493 (Resolved): teuthology-lock --status
- something similar to...
- 11:23 AM Feature #2413 (Resolved): qa: Test co-existence of sysvinit and upstart, 3: upstart controlled
- 09:39 AM Feature #2337: rgw and rados performance numbers
- Created a google doc with some testing plans. Check your mail!
- 09:36 AM CephFS Feature #2492 (New): ceph-fuse: improve ceph-fuse error reporting (especially missing cephx)
- If a client-side ceph.conf does not have "auth supported = cephx", but the server has it, ceph-fuse fails to mount, a...
- 05:22 AM Revision fa2f1dea (ceph): admin_socket: fix unit test
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:22 AM Revision ea1c0698 (ceph): mon: fix admin socket lock dependency
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:22 AM Revision 594f2bbd (ceph): perf_counters: use bufferlist instead of vector<char>
- bufferlist allocates page-sized chunks and avoids realloc, which'll be
more efficient than vector resize doubling jus... - 05:16 AM Revision 51de7c1a (ceph): osd: ack pending notify when unwatching
- If an unwatch crosses paths with a notify, we currently force the notifier
to keep waiting. Instead, implicitly ack ... - 04:58 AM Revision 5adc85bd (ceph): admin_socket: pass args separately
- This avoids making the callback parse off the command portion on their own.
It also lets them assert that the command... - 03:38 AM Revision e30e937c (ceph): admin_socket: register command prefixes
- Allow any command prefix to be registered, provided it is separated by a
space. Previously, we always matched agains... - 03:38 AM Revision ca8f3ba8 (ceph): admin_socket: streamline command set
- log flush
log dump
log reopen
config show
config set <var> <val>
perf dump
perf schema
Signed-off-by: Sage Wei... - 03:38 AM Revision 895e24d1 (ceph): admin_socket: streamline 'version' commands
- version -- ceph version
git_version -- git sha1
0 -- proto version (for backward compat)
Signed-off-... - 03:38 AM Revision 640eec9c (ceph): Merge remote branch 'gh/wip-admin'
- Reviewed-by: Yehuda Sadeh <yehuda@inktank.com>
05/30/2012
- 11:48 PM Revision c80bd9da (ceph): librbd: Simplify timing init
- Remove possibility of set_start_time before set_ictx error
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 11:22 PM Revision d8df1e99 (ceph): rgw: return default policy if policy attr is broken
- In case the policy attr on object/bucket is broken
return a default policy, in which the owner is set
as the bucket o... - 11:22 PM Revision 767664dc (ceph): rgw: put_bucket_info does not override attrs
- This fixes #2487. When writing bucket info we just
wrote the object content, and were overriding any
attrs that objec... - 11:18 PM Revision 04e66b42 (ceph): librbd: Add latency (elapsed-time) stats for rbd operations
- Fixes: #2408
Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
Reviewed-by: Jo... - 10:35 PM Revision a24145fc (ceph): log: add missing .cc file
- Signed-off-by: Sage Weil <sage@newdream.net>
- 10:28 PM Feature #2477: rados bench cleanup
- I think we need both:
* clean up objects after each run
* search+clean up objects from previous (maybe aborted) r... - 10:20 PM Bug #2491 (Resolved): watch/notify: racing notify and unwatch
- the following scenario (needs to be verified that it actually happens):
1. client 1 registers watch
2. client 2 reg... - 10:18 PM Feature #1240 (Resolved): teuthology: fail/clean up when daemons crash
- 10:17 PM CephFS Cleanup #814 (In Progress): hadoop: refactor hadoop shim in terms of java libceph bindings
- 10:16 PM Feature #2431 (Duplicate): teuthology: qemu + rbd testing
- 10:15 PM Revision 52187c3e (ceph): log: uninline SubsystemMap mutators
- This lets you do something like
p g_ceph_context->_conf->subsys.set_log_level(ceph_subsys_mon, 20)
from gdb.
Sign... - 10:15 PM Feature #2428 (Fix Under Review): auth: revise auth config params
- 10:15 PM Feature #2479 (Resolved): osd: runtime warning/message about syncfs variant in use
- 10:06 PM Revision a6f8fe35 (ceph): filestore: fix my broken ifdef flow
- Signed-off-by: Sage Weil <sage@newdream.net>
- 10:05 PM rgw Feature #2490 (New): rgw-admin: only register watch when needed
- Certain radosgw-admin operations don't need to use the cache notifications (e.g., log manipulation). We shouldn't reg...
- 09:54 PM Bug #2346 (Need More Info): xfs filesystem on top of rbd volume corrupts
- Has this issue been diagnosed?
- 09:41 PM Revision 188d1718 (ceph): doc: ceph.conf cleanup for cephx authentication.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 09:40 PM Revision 695f92d8 (ceph): doc: fix broken link.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 09:33 PM Feature #2463 (Resolved): adminsocket: 'show_config' command
- 09:29 PM Revision d05d98ce (ceph): doc: Added -a to service ceph stop.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 09:29 PM Revision cb5c6b5d (ceph): doc: Consolidated mkcephfs deployment files.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 09:28 PM Revision dd1d3880 (ceph): osd: fix null dereference
- The pg arg is optional!
Signed-off-by: Sage Weil <sage@inktank.com> - 08:48 PM Revision a3f4d595 (ceph): doc: Added pools and authentication.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 06:53 PM Revision a167213a (ceph): mon: fix reweight_by_utilization waiting
- Propose if we changed something, but not if we don't.
Signed-off-by: Sage Weil <sage@newdream.net> - 06:53 PM Revision 519fadbc (ceph): osd: fix rewewight_by_utilization
- Update the incremental, not the in-memory OSDMap!
Fixes: #2454
Signed-off-by: Sage Weil <sage@newdream.net> - 06:14 PM Revision 07498d66 (ceph): filestore: log about syncfs(2) etc support
- Fixes: #2479
Signed-off-by: Sage Weil <sage@newdream.net> - 06:02 PM Revision 2b856ace (ceph): Merge remote-tracking branch 'gh/stable'
- 05:14 PM Revision 6dd93b9a (ceph): doc: Updated authentication with header canonicalization.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 05:05 PM Revision e9588f0c (ceph): Merge remote-tracking branch 'gh/wip-2349'
- Reviewed-by: Sage Weil <sage@newdream.net>
- 04:55 PM Feature #2408: librbd: track latency with perfcounters
- Actually really
commit:c80bd9daea0c436fcf0af33fea9fc6b961279756 - 04:32 PM Feature #2408 (Resolved): librbd: track latency with perfcounters
- commit:04e66b420174d49e0f773cab1f6ef213dadc3397
- 04:30 PM Revision 26ecf608 (ceph): workloadgen: Fix the statistics output and a bug triggered by coll dest...
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 04:13 PM Revision f685f4d8 (ceph): filestore: initialize dummy xattr buffer
- Shut up valgrind.
Signed-off-by: Sage Weil <sage@newdream.net> - 03:51 PM Bug #2409 (Resolved): osd: pgs stuck in active
- 02:28 PM rbd Bug #2488 (Can't reproduce): rbd: EIO from rbd map
- ...
- 01:54 PM Bug #2487 (Resolved): rgw: (re)creating a suspended bucket succeeds
- When trying to create a bucket that already exists and is suspended, operation succeeds. Then the bucket appears in t...
- 11:51 AM Bug #2454 (Resolved): "rbd info xyz" hanging forever sometimes
- reweight-by-utilization was broken (modified in-memory OSDMap instead of updating pending incremental). fixed by com...
- 11:17 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Sage Weil wrote:
> for f in `seq 3 10`; do ceph osd reweight $f 1 ; done
Thank you very much.
I'm unable to prod... - 11:13 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- That's too bad.. i'll futz with it some here and see if i can reproduce.
You can fix your cluster with:
for f i... - 11:03 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Sage Weil wrote:
> Ok, the problem is that the weights are different on the osd's and mon's map. I'm not sure how t... - 10:55 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Ok, the problem is that the weights are different on the osd's and mon's map. I'm not sure how that could have happe...
- 10:16 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- progress:...
- 10:12 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Sage Weil wrote:
> Aha, ok I think we're on the right track. osd.2's map for 6924 says [3,15], but the client sent ... - 10:08 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Aha, ok I think we're on the right track. osd.2's map for 6924 says [3,15], but the client sent it to [2,15].
Can... - 10:03 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Sage Weil wrote:
> Just realized those were osd.3 messages mixed into the log, that's why it was confusing. However... - 09:56 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Just realized those were osd.3 messages mixed into the log, that's why it was confusing. However, osd.3 seems to thi...
- 09:37 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Sage Weil wrote:
>
> Is it possible to attach the whole osd2 log? Or a section that includes an osd map update. ... - 09:14 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Simon Frerichs wrote:
> Simon Frerichs wrote:
> > I've reattached the osd log because the first version hat some ol... - 08:22 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Simon Frerichs wrote:
> I've reattached the osd log because the first version hat some old data in it.
hum, lost ... - 08:16 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- I've reattached the osd log because the first version hat some old data in it.
- 08:07 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Here we go.
I've started the hanging rbd info process at 2012-05-30 16:46:35.281623 - 07:32 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Ok, the puzzling bit is that it isn't able to find the pg in question. It appears a bit hgher up in th elog:
May ... - 10:15 AM Feature #2486 (Resolved): crush: evaluate local retry behavior
- 10:14 AM Feature #2485 (Resolved): crush: validate hash function
- 10:04 AM Feature #2349 (Resolved): rados bench: Extra statistics
- 09:28 AM Bug #2484 (Resolved): ceph.spec does not list libuuid as build time dependency
- already fixed by commit:e868b44b3959a71c731f4ec9ff9773dead6dfcb5
- 08:33 AM Bug #2484 (Resolved): ceph.spec does not list libuuid as build time dependency
- See http://marc.info/?l=ceph-devel&m=133795298501327&w=2
I assume it's easy to add this to the .spec (or maybe it ... - 08:50 AM Bug #2481: ceph tell has almost no error reporting
- Greg Farnum wrote:
> This is actually a lot harder than you'd think — "ceph tell" works by sending a message to the ... - 08:14 AM Bug #2481: ceph tell has almost no error reporting
- This is actually a lot harder than you'd think — "ceph tell" works by sending a message to the monitors, which is the...
- 08:35 AM rbd Bug #2478: krbd: unmap on 3.4.0: scheduling while atomic...
- with patch above, this trace appears after ~15 sec from unmap operation
[ 2453.203365] ------------[ cut here ]---... - 08:28 AM CephFS Bug #2187: pjd chown/00.t failed test 97
- I haven't gone through the annoyance of counting it out, but it looks like the initially-reported test (112) is in th...
- 07:23 AM Linux kernel client Bug #2260: libceph: null pointer dereference at try_write+0x638+0xfb0
- I've been working for a week or two on refactoring the messaging
code in hopes to make it be based on a much cleaner... - 07:13 AM Linux kernel client Bug #2260: libceph: null pointer dereference at try_write+0x638+0xfb0
- Sorry I haven't updated this for some time.
In response to Yehuda: yes, I noticed that, and my instinct when
fir... - 07:10 AM Bug #2267 (In Progress): Ceph client crashed after shutting down one mds and osd
- > The variable which holds the NULL pointer is msg->bio_iter in]
> write_partial_msg_pages in net/ceph/messenger.c
... - 03:41 AM Revision 5842dcda (ceph): osd: fix misdirected op check
- The sense of this check was wrong from the get-go, back in a5d99add. Fix
it. This may be silencing a lot of misdire... - 01:34 AM Revision c5015ef3 (ceph): mon: add doxygen-format comments to the PaxosService class.
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:34 AM Revision 28a64fdb (ceph): mon: add doxygen-format comments to the Monitor class.
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:34 AM Revision 1816856e (ceph): mon: Do some aesthetic reformatting on the Paxos class.
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:34 AM Revision 8c95e8bc (ceph): mon: add doxygen-format comments to the Elector class.
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:34 AM Revision 15a30d92 (ceph): mon: Do some aesthetic reformatting on the Elector class.
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:34 AM Revision 05469b2b (ceph): mon: add some info on can_mark_out function
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:34 AM Revision 504c6ce9 (ceph): doc: Add Paxos call chain to doc/mon/
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 01:34 AM Revision 7064a9c0 (ceph): mon: add doxygen-format comments to the Paxos class.
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 12:09 AM CephFS Bug #1047: mds: crash on anchor table query
- We are auto creating up to 200 preconfigured home directories on all four nodes, each home dir consists of ca. 400 di...
05/29/2012
- 11:52 PM Revision 92fec423 (ceph): test_filestore_workloadgen: write zeros, not random bytes
- The random byte generation is sufficiently slow that it skews the results.
Zeros are good enough!
Signed-off-by: Sag... - 11:52 PM Revision b46b815a (ceph): test_filestore_workloadgen: less noise
- Signed-off-by: Sage Weil <sage@newdream.net>
- 11:31 PM Revision 4e2846b4 (ceph): config: add 'set_config <var> <val>' admin socket command
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:31 PM Revision 29e6f8c5 (ceph): log: add log flush, dump_recent, reopen admin socket commands
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:31 PM Revision bb0c7c6a (ceph): admin_socket: fix unittest
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:57 PM Bug #2454: "rbd info xyz" hanging forever sometimes
- Found a copy of the old osdmap at one monitor. I've attached it.
- 09:40 PM Bug #2454: "rbd info xyz" hanging forever sometimes
- Well, i see one (unrelated) oddity in this code, but it doesn't explain how the non-existent message is coming up. (...
- 11:23 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- if you need more information, feel free to ping me.
- 09:07 PM Bug #2409: osd: pgs stuck in active
- oh, bother. this is wip-osd-peering not getting backported to dho branch. commit:17114f266a336b6edd7e98975d494fdd48...
- 12:41 PM Bug #2409: osd: pgs stuck in active
- saw this on burnupi. cranked up logging. hopefully we'll catch it again...
- 08:33 PM Revision 6849f82e (ceph): rados.cc: added listomap and getomap commands
- Signed-off-by: Eleanor Cawthon <eleanor.cawthon@inktank.com>
- 07:07 PM Revision ab8f6b1b (ceph): admin_socket: initialize explicitly on startup; disallow changes
- There is an annoying dependency between the config lock and the admin
socket lock due to the fact that we initialize ... - 07:07 PM Revision a3f81376 (ceph): config: add 'show_config' admin socket command
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:23 PM Revision f953c4c0 (ceph): librados.h: update rados_object_list_next documentation for key/locator
- *entry and *key don't need to be freed, and are only valid until the
next call to rados_objects_list_*.
Signed-off-b... - 05:10 PM Feature #2349 (Fix Under Review): rados bench: Extra statistics
- 05:09 PM Revision 1277a465 (ceph): cephcontext: generalize admin socket hook
- We'll process non-perfcounter commands too.
Signed-off-by: Sage Weil <sage@inktank.com> - 03:20 PM rgw Documentation #2483 (Resolved): doc: radosgw api diffs to swift
- 03:19 PM rgw Documentation #1813 (Resolved): doc: document radosgw api diffs with s3
- 03:08 PM rgw Bug #2482 (Rejected): rgw: duplicate content-length results in 400
- Not sure if it's from apache of haproxy. While this is not really a bug, we may want to change this behavior.
- 02:20 PM Bug #2481 (Won't Fix): ceph tell has almost no error reporting
- fumbling around as a noob, I tried all of:
ubuntu@burnupi50:/etc/ceph$ ceph osd tell 1 injectargs 'debug osd = 0'
... - 01:43 PM Documentation #155 (In Progress): document ceph auth
- 01:41 PM Documentation #2474 (Resolved): re-document using autobuilt branches
- Needed to restore from git, update the hyperlinks, and add it into the package section. Verified with Sage.
- 01:30 PM Feature #2412 (Resolved): qa: Test co-existence of sysvinit and upstart, 2: sysvinit controlled, ...
- Changing state to reflect that the case has been tested and passed.
- 01:29 PM Feature #2411 (Resolved): qa: Test co-existence of sysvinit and upstart, 1: sysvinit controlled, ...
- Updating status to reflect that the fix has been tested and passed. "Resolved" is the closest option for this.
- 12:41 PM Bug #2464 (Resolved): osdmap: assert in get_inst()
- 09:13 AM Bug #2480: mon: election won't complete
- probable regression from mon bootstrap changes.
- 09:11 AM Bug #2480 (Rejected): mon: election won't complete
- ...
- 08:52 AM CephFS Bug #2187: pjd chown/00.t failed test 97
- ceph-fuse *and* kclient affected.. mds?...
- 08:47 AM Feature #2479 (Resolved): osd: runtime warning/message about syncfs variant in use
- 04:02 AM Bug #2267: Ceph client crashed after shutting down one mds and osd
- The variable which holds the NULL pointer is msg->bio_iter in write_partial_msg_pages in net/ceph/messenger.c. I conf...
- 12:47 AM CephFS Bug #1047: mds: crash on anchor table query
- This bug seems to be reproduced with the deadlock described on the mailing list as
"OSD deadlock with cephfs client ...
05/26/2012
- 01:30 PM rbd Bug #2478 (In Progress): krbd: unmap on 3.4.0: scheduling while atomic...
- commit:cd9d9f5df6098c50726200d4185e9e8da32785b3 already fixes this; we just neglected to send it upstream for 3.4. I...
- 06:05 AM rbd Bug #2478 (Can't reproduce): krbd: unmap on 3.4.0: scheduling while atomic...
- On 3.4.0 on the 'rbd unmap' call kernel rbd driver produces following trace(same for vanilla):
[ 8660.786269] BUG... - 09:10 AM CephFS Bug #2385: max mds = 2, mds hang and crash
- net.core.wmem_max = 536870912
net.core.rmem_max = 536870912
net.core.wmem_default = 262144
net.core.rmem_default =...
05/25/2012
- 11:15 PM Revision 73e29b65 (ceph): doc: needed carriage returns between headings, code tag and syntax.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 11:03 PM Revision d33f4fb6 (ceph): doc: removed redundant install.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 11:03 PM Revision 68874afa (ceph): doc: Filled in descriptions for object operations.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 04:29 PM Revision e868b44b (ceph): ceph.spec.in: buildrequires libuuid-devel
- Reported-by: Guido Winkelmann <guido-ceph@thisisnotatest.de>
Signed-off-by: Sage Weil <sage@inktank.com> - 04:28 PM Revision 5eda9b70 (ceph): ceph.spec.in: drop unnecessary rmdirs
- This hunk was inadvertantly included in a330c64.
Reported-by: Guido Winkelmann <guido-ceph@thisisnotatest.de>
Signed... - 02:27 PM Feature #2412: qa: Test co-existence of sysvinit and upstart, 2: sysvinit controlled, using /var/...
- sudo service ceph start -pass
sudo /etc/init.d/ceph start - pass
reboot. - pass
sudo initctl list|grep ceph" has o... - 02:24 PM Feature #2411: qa: Test co-existence of sysvinit and upstart, 1: sysvinit controlled, outside of ...
Tested
sudo service ceph start - pass
sudo /etc/init.d/ceph start - pass
reboot. - pass
sudo initctllist|gre...- 09:54 AM Feature #2472 (In Progress): osd: add opaque 'class <name> <foo>' cap that class can interpret/en...
- 09:54 AM Feature #2471 (Fix Under Review): osd: add prefix match to OSDCaps
- "allow rwx pool=rbd prefix rbd_data; allow x pool=rbd prefix rbd_header"
"allow rwx pool=rbd; deny rw pool=rbd prefi... - 04:05 AM CephFS Bug #1047: mds: crash on anchor table query
- Just hit this bug again with 0.47.2 on ext4, so it is still open and it is still an issue.
05/24/2012
- 11:54 PM Revision 6601ef95 (ceph): doc: Restored install for dev/unstable packages.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 10:01 PM Revision ba540623 (ceph): filestore: make SNAP_DESTROY ioctl check more informative
- Tell user about user_subvol_rm_allowed mount option.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:53 PM Revision 07c9eda2 (ceph): filestore: fix btrfs subvol ioctl tests
- Create a temp subvol that *we own* to test snaps against. Otherwise, we
won't be able to remove them if we weren't t... - 09:07 PM Revision bea1e031 (ceph): filestore: verify that op_seq is nonzero
- If we get zero from commit_op_seq something is very wrong, and we should
not continue. Broken mkfs or something dele... - 09:05 PM Revision 18b0d657 (ceph): vstart.sh: wipe out btrfs subvols in osd dirs
- For testing on btrfs.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:05 PM Revision 0f38d756 (ceph): filestore: fix initial btrfs snap creation (on mkfs)
- Create the snap_1 snap of current *after* we write the op_seq. Otherwise
we will mistakingly believe the snapshot is... - 08:07 PM Revision 3ae84eec (ceph): rgw: replace dump_format() with dump_int()
- Cleanup, fixes issue #2469.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 04:47 PM Revision ca79f45a (ceph): Merge branch 'stable'
- 04:45 PM Revision 9a9418eb (ceph): librados, ObjectCacher: include limits.h for INT_MAX
- This was required to build on CentOS 6.2. I'm not sure why it worked
without it on Debian.
Signed-off-by: Josh Durgi... - 04:37 PM Revision ea6c546e (ceph): Added python-dev to list of required packages.
- 01:55 PM rgw Cleanup #2469 (Resolved): rgw: replace Formatter->dump_format(..., "%d", ...) with Formatter->dum...
- Fixed, commit:3ae84eec2dbcb83a31924cc348e9b91157ecb3b1.
- 11:02 AM Bug #2475: rbd.py can leave Image object in inconsistent state on failure to construct
- In the interactive interpreter, "the next statement" is where the interpreter knows the prior scope is closed, and de...
- 10:35 AM Bug #2475 (Resolved): rbd.py can leave Image object in inconsistent state on failure to construct
- commit:8c1c882999b8eef2dbf4fca066a5978e547db83e
- 10:54 AM Feature #2477 (Resolved): rados bench cleanup
- Right now rados bench write leaves a lot of data just lying around the cluster, without a good way to clean itself up...
- 09:16 AM Bug #2476 (Resolved): osd: watch timeout depends on operations to an object
- The watch timeout is an in-memory thing that's local to the primary. If the primary changes, the timer for ending the...
- 09:16 AM rgw Bug #2465 (Resolved): rgw: bad marker output when listing a bucket
- Fixed, commit:9094dce4cc232fc70fbc6d4789a79dd6af6ff4b6.
- 05:43 AM CephFS Bug #733: cmds crash: mds/LogEvent.cc:88: FAILED assert(p.end())
- here you go. a log with ms = 1 and mds = 20.
dirs and files are replaced with 'o's. - 02:35 AM Revision 8c1c8829 (ceph): rbd.py: Fix seemingly-random segfaults
- Avoid resetting Image.closed until image is successfully opened.
Fixes: #2475
Signed-off-by: Dan Mick <dan.mick@inkt... - 02:33 AM Revision 2f963fa7 (ceph): librados: Remove errant sha1
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com> - 02:33 AM Revision 6e46de37 (ceph): rbd.py doc: Fix markup on :class: references
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
05/23/2012
- 11:42 PM Revision 9453413c (ceph): Merge branch 'osd-mkfs'
- Reviewed-by: Sam Just <sam.just@inktank.com>
- 11:41 PM Revision b5e7fdd3 (ceph): Merge remote-tracking branch 'gh/wip-mon-auth'
- Reviewed-by: Yehuda Sadeh <yehuda@inktank.com>
- 11:41 PM Revision fbae6031 (ceph): Merge branch 'stable'
- 11:31 PM Revision c18e4d80 (ceph): filestore: mkfs: do not clobber exisiting journal if it has matching fsid
- If the journal is for *this* fs, do not clobber. If it appears invalid in
any way, initialize it.
Signed-off-by: Sa... - 11:31 PM Revision a980d018 (ceph): osd: do not clobber existing key with --mkkey
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:31 PM Revision 5bf5e91f (ceph): osd: require cluster fsid for mkfs
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:31 PM Revision c96aa62d (ceph): osd: verify fsid whoami to match on repeat mkfs
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:31 PM Revision 386139bb (ceph): filestore: more informative fsid mkfs debug output
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:31 PM Revision 7805978c (ceph): filestore: make read_fsid() out arg be a pointer
- As per style guide.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:31 PM Revision ec7bf564 (ceph): filestore: fix misplaced #endif
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:31 PM Revision 598dea12 (ceph): filestore: mkfs: only create snap_0 if we created current_op_seq
- We only need to create snap_0 if we just created current_op_seq and this is
a brand-new dir.
Signed-off-by: Sage Wei... - 11:31 PM Revision b04b73d4 (ceph): filestore: mkfs: less noise about leveldb creation
- and make it clear that we didn't necessarily create it; it may have already
been there.
Signed-off-by: Sage Weil <sa... - 11:31 PM Revision c498827f (ceph): filestore: drop useless mkjournal error check
- We don't care!
Signed-off-by: Sage Weil <sage@inktank.com> - 11:31 PM Revision a074d0b5 (ceph): filejournal: add check() method
- Check if a journal appears to be valid, where valid means the header is
intact and has a matching fsid.
Signed-off-b... - 11:12 PM Revision 09192d65 (ceph): doc/api/s3/bucketops.rst Corrected heading.
- doc/install/chef.rst Added Sam's edits to the Chef install.
Signed-off-by: John <john@admin-host.(none)> - 11:09 PM Revision bf2028fc (ceph): rgw: replace all bad use of formatter->dump_format()
- This fixes #2465. We should never ever use the dumped value
as the format. Never.
Signed-off-by: Yehuda Sadeh <yehud... - 10:44 PM Revision 3956cb8f (ceph): doc/api/s3 <-chipping away at objectops and clean up. Added Copy Object.
- doc/source <-fixed a typo, updates for installing on precise, rm $ from cmds
doc/start <-replaced clone link with ta... - 09:03 PM Revision 36951d6e (ceph): mon: use temp access_{r,all,cmd} variables to simplify the checks
- Avoid duplicating the compound check all over the place.
Signed-off-by: Sage Weil <sage@inktank.com> - 09:02 PM Revision 176a14ae (ceph): debian, upstart, osd: osd disk preparation and activation scripts
- Signed-off-by: Tommi Virtanen <tv@inktank.com>
- 09:02 PM Revision f1e4d44d (ceph): upstart: support mds
- Closes: #2414
Signed-off-by: Tommi Virtanen <tv@inktank.com> - 08:49 PM Revision 18af8137 (ceph): mon: drop redundant get_allow_all() cap check
- check_privileges() already does this.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:24 PM Revision 1870e12c (ceph): msgr: add mistakenly-removed msgr->lock.Lock() back in
- This was accidentally deleted in
828f619c84c02edec10ce0c22d47a72136ba3749 but is definitely required.
Signed-off-by:... - 07:46 PM Revision 475e07a2 (ceph): debian: Create placeholder dirs for mon and osd data directories.
- Signed-off-by: Tommi Virtanen <tv@inktank.com>
- 07:05 PM Bug #2475 (Resolved): rbd.py can leave Image object in inconsistent state on failure to construct
- Constructing an Image object with bad parameters (say, an nonexistent rbd image name) can leave the
resulting object... - 06:25 PM Revision ab04d95a (ceph): monmap: return ENOENT from build_from_host_list when no mons
- Make sure we return an error when no monitors are parsed.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:25 PM Revision 80c6278f (ceph): monmap: ignore conf mon entries when -m is specified
- Use -m <addrs> instead of conf mons (instead of combining them into a
single large search pool).
Signed-off-by: Sage... - 05:55 PM Revision cf6db40a (ceph): Makefile.am: only append libs3 to SUBDIRS when needed
- We appended libs3 to SUBDIRS unconditionally, we only need that
when we actually want to compile it.
Signed-off-by: ... - 05:55 PM Revision 244992df (ceph): libs3: remove make install target
- 05:55 PM Revision a330c64c (ceph): Update ceph.spec for ceph-0.47
- Add BuildRequires: libxml2-devel.
Move BuildRequires: libcurl-devel to a more proper place.
[libs3 and librgw.h bits... - 05:55 PM Revision 26843ada (ceph): Makefile: do not install librgw.h
- Signed-off-by: Sage Weil <sage@newdream.net>
- 05:55 PM Revision 8bf9fde8 (ceph): v0.47.2
- 05:48 PM Documentation #2474 (Resolved): re-document using autobuilt branches
- It seems John removed the docs for setting up autobuilt apt repos, the signing key etc. That's still needed.
- 05:40 PM Revision f461e132 (ceph): monclient: sanity check on monmap size
- Signed-off-by: Sage Weil <sage@newdream.net>
- 05:39 PM Revision d173c7bf (ceph): monmap: avoid dup addrs in generated initial monmap
- We were getting a dup addr from -m <foo> and ceph.conf. Prefer the
ceph.conf name in this case.
Also check for dup ... - 05:38 PM Revision 75e9c0cb (ceph): monmap: aseert addrs are unique on add(), calc_ranks()
- Signed-off-by: Sage Weil <sage@newdream.net>
- 04:56 PM rgw Feature #2473 (Resolved): rgw: revisit operation logging
- Sending append for each client operation is expensive. We can definitely find better solution.
- 04:49 PM CephFS Bug #733: cmds crash: mds/LogEvent.cc:88: FAILED assert(p.end())
- We'll need a detailed log (and possibly access to the data that's causing the crash) to diagnose this. Can you turn o...
- 02:25 PM CephFS Bug #733: cmds crash: mds/LogEvent.cc:88: FAILED assert(p.end())
- here is a backtrace:
Core was generated by `/usr/bin/ceph-mds -i alpha --pid-file /var/run/ceph/mds.alpha.pid -c ... - 10:55 AM CephFS Bug #733: cmds crash: mds/LogEvent.cc:88: FAILED assert(p.end())
- I get the same with v0.47.1:
0> 2012-05-23 19:50:20.105956 7f7c87482700 -1 mds/LogEvent.cc: In function 'stat... - 04:41 PM Feature #2405 (Resolved): osd: Make ceph-osd --mkfs idempotent
- 04:41 PM Bug #2443 (Resolved): Anyone can list all keys, even with caps mon 'allow rwx' and not 'allow *'
- commit:b5e7fdd3d7a413e03bb5bb43b689e06c9cd6ffd9
- 04:20 PM Bug #2470 (Closed): cookbook: keyword "relase" in apt.rb causes wget to fail.
- Fixed in ab51f4dcd69774411015548db46dc18c198e4181.
- 03:15 PM Bug #2470 (Closed): cookbook: keyword "relase" in apt.rb causes wget to fail.
- in ceph-cookbooks/ceph/recipes: the keyword "relase" should be "release". The wget fails in it's current form but co...
- 04:08 PM Feature #2413: qa: Test co-existence of sysvinit and upstart, 3: upstart controlled
- Branch to test is now available as "chef-3".
- 04:08 PM Feature #2412: qa: Test co-existence of sysvinit and upstart, 2: sysvinit controlled, using /var/...
- Branch to test is now available as "chef-3".
- 04:08 PM Feature #2411: qa: Test co-existence of sysvinit and upstart, 1: sysvinit controlled, outside of ...
- Branch to test is now available as "chef-3".
- 04:05 PM Feature #2472: osd: add opaque 'class <name> <foo>' cap that class can interpret/enforce
- allow class 'foo' bar
Should allow the class 'foo' to do the operation IF it is happy with 'bar'. That is, the cl... - 04:02 PM Feature #2472 (New): osd: add opaque 'class <name> <foo>' cap that class can interpret/enforce
- 04:03 PM Feature #2471: osd: add prefix match to OSDCaps
- "allow prefix 'foo' r" should allow 'r' access only to objects (and locator keys) that are prefixed by 'foo'.
- 04:02 PM Feature #2471 (Resolved): osd: add prefix match to OSDCaps
- 02:17 PM Documentation #2271 (In Progress): FAQ: BTRFS vs XFS
- http://ceph.com/docs/master/rec/filesystem/ still needs some info on how easy it is to change from XFS to btrfs.
- 12:59 PM Feature #2399 (Resolved): qa: haproxy + rgw + jenkins
- 12:57 PM rgw Cleanup #2469 (Resolved): rgw: replace Formatter->dump_format(..., "%d", ...) with Formatter->dum...
- 12:52 PM Feature #2426 (Resolved): precise packages for apache2, fastcgi
- OK, I think the packages are workable now.
- 12:52 PM rbd Feature #2468 (Resolved): librbd: provide a way for a user to flush and invalidate the cache
- This could be an admin socket command. This would make live migration work with older qemu when rbd caching is enabled.
- 12:37 PM rbd Feature #2467 (Resolved): qemu: implement bdrv_invalidate_cache
- This is used during live migration to clear librbd's cache on the destination host before starting the guest there. I...
- 12:35 PM rbd Feature #2466 (Resolved): librbd: add invalidate_cache function to interface
- Qemu requires this to make live migration work when caching is enabled.
- 12:10 PM Bug #2219: OSD's commit suicide with 0.44
- i added one node to my small test cluster. it has now totally three nodes. so rados is filling the new node.
the mac... - 11:29 AM Bug #2219: OSD's commit suicide with 0.44
- Eric Dold wrote:
> dmesg looks ok to me.
was the system heavily loaded?
do you have a core file?
this basic... - 11:09 AM Bug #2219: OSD's commit suicide with 0.44
- dmesg looks ok to me.
- 09:05 AM Bug #2219: OSD's commit suicide with 0.44
- Eric Dold wrote:
> I just hit this with v0.47.1:
>
> 2012-05-23 13:33:08.958564 7fe61124d700 -1 common/Heartbeat... - 05:31 AM Bug #2219: OSD's commit suicide with 0.44
- I just hit this with v0.47.1:
2012-05-23 13:33:08.958564 7fe61124d700 -1 common/HeartbeatMap.cc: In function 'boo... - 11:21 AM rgw Bug #2465: rgw: bad marker output when listing a bucket
- Actually, it'll only happen with '%'. We were using formatter->dump_format() istead of formatter->dump_string().
- 11:03 AM rgw Bug #2465 (Resolved): rgw: bad marker output when listing a bucket
- When providing a marker that contains '%' (and possibly other characters that are url-escaped), the returned result c...
- 09:32 AM Bug #2464 (Resolved): osdmap: assert in get_inst()
- INFO:teuthology.orchestra.run.err: ceph version 0.47.1-128-gc31ab04 (commit:c31ab04b572ef5b5cf622b33b92c9e17c7070662)...
- 09:17 AM Linux kernel client Bug #2389: rbd: hung xfstest 67
- ubuntu@teuthology:/a/nightly_coverage_2012-05-22-b/3062
- 09:16 AM CephFS Bug #2187: pjd chown/00.t failed test 97
- 2012-05-22T12:08:06.355 INFO:teuthology.task.workunit.client.0.out:Test Summary Report
2012-05-22T12:08:06.355 INFO:... - 07:56 AM Feature #2463 (Resolved): adminsocket: 'show_config' command
- 03:47 AM Revision c31ab04b (ceph): doc: librados: fix getxattr return code
- Fixes: #2391
Signed-off-by: Sage Weil <sage@inktank.com> - 01:50 AM Bug #2462 (Resolved): osd/PG.cc: 402: FAILED assert(log.head >= olog.tail && olog.head >= log.tail)
- 2012-05-23 06:16:37.080317 7f18f6012700 -1 osd/PG.cc: In function 'void PG::merge_log(ObjectStore::Transaction&, pg_...
- 01:31 AM Revision fb6ed150 (ceph): mon: apply 'r' and 'w' caps to mdsmap
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:31 AM Revision e895d61f (ceph): mon: apply 'r' and 'w' caps to mon and pg commands
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:31 AM Revision 311a23ce (ceph): mon: restrict mon 'join' message/command
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:31 AM Revision c43c7744 (ceph): mon: require admin privs to issue any monitor commands
- This is overkill, but a first step before pushing caps enforcement down
into each subfunction that processes monitor ... - 01:31 AM Revision 7d0552a1 (ceph): mon: push down command auth checks to each command
- Everything except 'fsid', which I figure is truly public information?
Signed-off-by: Sage Weil <sage@inktank.com> - 01:31 AM Revision def15d7f (ceph): mon: push down 'auth ...' command auth checks
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:31 AM Revision b7f27917 (ceph): mon: fix error messages
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:31 AM Revision 880e86b2 (ceph): mon: allow health, status, log, etc. messages with 'r' cap
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:31 AM Revision c1c93e57 (ceph): mon: apply 'r' and 'w' caps to osdmap
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:31 AM Revision 2dec922f (ceph): vstart.sh: set HOSTNAME when -l is used
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:30 AM Revision 87014391 (ceph): mon: fix 'get-or-create-key'
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:24 AM Revision 17c4624e (ceph): osd: only (re)write_meta files if old content is incorrect
- Do nothing if the file already exists with the correct content.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:24 AM Revision 05db988a (ceph): filestore: remove unused 'filestore dev' option
- Eons ago we would call mount/unmount on this, but it's been dead for a
while now.
Signed-off-by: Sage Weil <sage@ink... - 12:24 AM Revision 58d8fafb (ceph): filestore: do not wipe on mkfs
- Do not wipe out old content in the data dir.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:24 AM Revision 143762f5 (ceph): filestore: always open/verify basedir; clean up error messages
- Verify basedir exists immediately.
Clean up error output.
Clean up label names.
Signed-off-by: Sage Weil <sage@inkta... - 12:24 AM Revision 36503d1b (ceph): filestore: mkfs: rename buf -> fsid_fn
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:24 AM Revision d923e338 (ceph): filestore: pass target into read_fsid()
- Don't assume it should be read into this->fsid.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:24 AM Revision 79d60770 (ceph): filestore: mkfs: only (re)set fsid if not already set
- Only set fsid if it is not set.
Also, error out if it exists and does not match.
Signed-off-by: Sage Weil <sage@ink... - 12:24 AM Revision b7a85a8f (ceph): filestore: create but do not clobber current/commit_op_seq
- If it exists, preserve its value. Otherwise, create and set it to 1.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:24 AM Revision b01dbcdd (ceph): filestore: leave existing current/
- If current exists, leave it be, but:
* make sure it's a dir
* check if it's a btrfs snapshot, and set btrfs_stable... - 12:24 AM Revision 53b05d58 (ceph): osd: only create superblock if not present
- If it exists, leave it be.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:13 AM Revision 55c21de5 (ceph): Merge branch 'wip-quorum'
- 12:10 AM Revision ff64818f (ceph): doc/config-cluster/chef.rst <-fixed an error in upload command.
- doc/config-cluster/deploying-ceph-conf.rst <-added sudo. required.
doc/config-cluster/deploying-ceph-with-mkcephfs.rs...
05/22/2012
- 10:03 PM Bug #2461 (Resolved): DBObjectMap is incompatible with collection_rename
- Objects are stored using a (collection_name, object_name) prefix. When a collection is renamed from A to B, objects ...
- 09:41 PM Bug #2234 (Resolved): Sometimes 'ceph -s' is unable to show pg data and crashes
- This code has all been replaced!
- 09:40 PM Bug #2391 (Resolved): librados docs bug
- commit:c31ab04b572ef5b5cf622b33b92c9e17c7070662
- 09:34 PM Bug #2379: Mon crash after start
- If this happens again, can you grab a tarball of the mon data directory before fixing/restarting?
Also, if you cou... - 09:24 PM Bug #2443 (Fix Under Review): Anyone can list all keys, even with caps mon 'allow rwx' and not 'a...
- see wip-mon-auth
- 08:53 PM Revision c4447230 (ceph): Merge remote-tracking branch 'origin/master' into wip-msgr-interface
- Conflicts:
src/msg/Messenger.h
src/msg/SimpleMessenger.h
Signed-off-by: Greg Farnum <greg@inktank.com> - 08:53 PM Revision 53269374 (ceph): config: add --show-args option
- The --show-args dumps the arguments passed on the command line.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 07:54 PM Revision 0ae59eae (ceph): Merge branch 'stable'
- 07:48 PM Revision 11030793 (ceph): objectcacher: fix infinite loop in flusher_entry
- The addition of accounting for simultaneous writers in
f3043fee3e22600cb4349072287842db129588eb could lead to an infi... - 07:33 PM Feature #2426: precise packages for apache2, fastcgi
- Iiuc its apt and not dpkg that checks sigs. Creating and signing a repo and pointing apt at it is enough to generate...
- 07:27 PM Feature #2426: precise packages for apache2, fastcgi
- So I've built the packages from the git repo versions (not source, but binary) and they install on my
precise deskto... - 05:22 PM Feature #2405 (Fix Under Review): osd: Make ceph-osd --mkfs idempotent
- 05:22 PM Feature #2418 (Resolved): mon: Take mandatory initial quorum members from ceph.conf
- 05:22 PM Feature #2419 (Resolved): mon: Take peer hints via admin socket
- 05:06 PM Feature #2427 (Resolved): precise gitbuilder http
- 10:38 AM Feature #2427: precise gitbuilder http
- Fixed commit:0e4f131ebfd0dd43593d1d95b544318fb749ad53
lighttpd is now installed and configured - 04:55 PM rgw Bug #2027 (Can't reproduce): rgw -> apache miscommunication
- 04:54 PM Revision 884aa8a6 (ceph): Merge branch 'master' of github.com:ceph/ceph
- 04:53 PM Revision b0d6895a (ceph): doc/config-cluster/ceph-conf.rst <-Fixed some formatting issues.
- doc/config-cluster/deploying-ceph-conf.rst <-minor edit.
doc/start/get-involved-in-the-ceph-community.rst <-removed $... - 04:50 PM rgw Feature #1726 (Rejected): rgw: improve multipart upload performance
- Actually, this is obsolete. Nowadays when we complete the upload we only create a new index, which we point at all th...
- 02:05 PM Revision d7a8084b (ceph): monmap: calc_ranks() on rename()
- This is simpler than ensuring we track all the right invariants.
Signed-off-by: Sage Weil <sage@inktank.com> - 12:45 PM Bug #2454: "rbd info xyz" hanging forever sometimes
- I notice that the pgid output on these lines don't match, even though they're using the same output function and the ...
- 11:40 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Some additional information:
I'm running Kernel 3.3.6 from kernel.org compiled two days ago from kernel.org on the t... - 11:27 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- I had some spare time to do some testing, today.
As mentioned on my initial post we're running KVM VPS on Ceph.
I r... - 11:01 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Oh geeze, my bad!
- 10:50 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Greg, this has nothing to do with the kernel - it's the rbd command line tool. I can't seem to change the tracker bac...
- 10:39 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- Argh, you're right. This looks like a problem with a CRUSH mismatch between the userspace and kernelspace implementat...
- 12:36 AM Bug #2454: "rbd info xyz" hanging forever sometimes
- I've attached rbd and osd log. The hanging rbd was started at 09:10:33 and ended at 09:12:18.
One thing came to my... - 12:02 PM rgw Feature #2460 (Rejected): rgw: support multiple ceph backends
- 11:58 AM Feature #2459 (Rejected): admin socket: config reload
- Currently, the only way to reload configuration is by restarting daemons.
- 09:27 AM Feature #2295 (Resolved): make qemu cache=writeback,writethrough option turn on librbd caching
- Applied upstream, should be in qemu 1.2.
- 08:18 AM Bug #2446 (Resolved): libceph: corrupt inc osdmap epoch 24630 off 702 (ffff88001e5d876c of ffff88...
- Thanks for testing!
- 08:04 AM Bug #2446: libceph: corrupt inc osdmap epoch 24630 off 702 (ffff88001e5d876c of ffff88001e5d84ae-...
- I have tested this patch for a couple hours today and there were no 'corrupt inc osdmap' messages. Thanks.
- 03:01 AM Revision 4fd52491 (ceph): Merge branch 'wip-defensive-alloc'
- 03:00 AM Revision cfe261bf (ceph): crush: check reweight return value
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:36 AM Revision 5959ea7b (ceph): doc: fix broken link in FAQ
- Signed-off-by: Ross Turk <ross@inktank.com>
- 01:22 AM Revision cf584867 (ceph): doc: added links from the chef install doc to the chef config doc and back
- Signed-off-by: Ross Turk <ross@inktank.com>
- 01:21 AM Revision 6149fb70 (ceph): doc: Added a FAQ page.
- Added a FAQ page, which we will need to put more content in.
Signed-off-by: Ross Turk <ross@inktank.com> - 01:01 AM Revision 828f619c (ceph): msgr: pass a Connection instead of a Pipe through submit_message.
- This lets us maintain a user's existing Connections even if the underlying
Pipe object disappears (ie, it's lossy and... - 12:49 AM Revision e3b3108b (ceph): Added the Swift API as is. Still needs a lot of work.
- Signed-off-by: John Wilkins <john.wilkins@dreamhost.com>
- 12:24 AM Revision de7fb970 (ceph): obj_bencher: add --show-time option for dumping time
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
05/21/2012
- 11:57 PM Revision d93410c7 (ceph): doc: document range of methods for cluster creation, expansion
- This isn't quite complete, and should be integrated with the info in the
ops section on mon cluster expansion, but th... - 11:43 PM Revision c9e98962 (ceph): obj_bencher: adding stddev to bandwidth and latency
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 11:39 PM Revision 812989bf (ceph): doc: misc updates
- doc/architecture.rst - removed broken reference.
doc/config-cluster - cleanup and added chef
doc/install - Made gener... - 11:13 PM Revision bc9e592b (ceph): obj_bencher: add min/max bandwidth
- Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
- 10:07 PM Revision 133cd69a (ceph): mon: fix 'no initial monitors' warning
- It is valid to start with no initial monitors even when the initial set is
not defined; it just means that we are onl... - 09:32 PM Revision 3a2dc969 (ceph): libs3: remove make install target
- 09:28 PM Revision f5a94044 (ceph): v0.47.1
- 09:27 PM Revision 4e3807bc (ceph): Makefile.am: only append libs3 to SUBDIRS when needed
- We appended libs3 to SUBDIRS unconditionally, we only need that
when we actually want to compile it.
Signed-off-by: ... - 09:21 PM Revision 13b350a5 (ceph): filestore: test xattrs on temporary file
- There isn't one we can rely on existing yet. Broken by
1314a00798ed4b7ef2f2686f0195c5c53c98c2ce while fixing #2452.
... - 09:20 PM Revision 6c0a1220 (ceph): ceph-object-corpus: add some old v2 monmaps
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:59 PM Cleanup #2458 (Resolved): filestore: backend abstraction
- We should create a backend abstraction layer in the filestore. This layer will hold all the filesystem specific opera...
- 08:58 PM Revision c06fb20a (ceph): monmap: fix decoding of old monmaps
- The old monmaps had a u16 for the version. Fix our decode.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:15 PM Revision a4d034d0 (ceph): qa: fix mon single_host_multi.sh test
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:15 PM Revision 731effba (ceph): mon: rename ourselves if our name does not match committed monmap
- If we are part of the quorum but our name does not match, fix it.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:14 PM Revision cd9d8622 (ceph): mon: only rename peers in seed monmap
- Once we have a committed monmap, stop renaming peers. This only gets us
confused because our monmap does not match t... - 08:07 PM Revision e1564aff (ceph): mon: move election msg check into elector
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:01 PM Revision ee138678 (ceph): mon: allow mons to rename themselves with the join message
- This allows mons that got stuck in the initial quorum's monmap as noname-*
to fix their name before joining in.
Sign... - 08:00 PM Revision b4d2b1e4 (ceph): monmap: fix rename
- Update addr_name map appropriately.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:46 PM Revision 29a158a6 (ceph): mon: drop useless rank racalc in init
- bootstrap() handles this, no need to worry about it here.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:38 PM Revision dae2f532 (ceph): mon: fix leak of MonMap
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:36 PM Revision 7b2614b5 (ceph): monmap: filter_initial_members -> set_initial_members
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:40 PM Revision 13c86ca2 (ceph): Update leveldb to remove CompactionInputErrorParanoid
- CompactionInputErrorParanoid tends not to complete in time during
gitbuilder runs resulting in false failures.
Signe... - 04:24 PM Bug #2454: "rbd info xyz" hanging forever sometimes
- Yep, the OSD gets the message, acks it, and acks several following pings. Looks like some strange bug on the OSD side...
- 03:56 PM Revision 7d5fa4e5 (ceph): man: Remove leftover --bin from ceph-authtool(8).
- Regenerate manpage.
See 4832c46eab55b1f2cbe9aca419124c2c9d645889,
ec732816ef04e036db338cdbe5cdfb9a1b649f8e for the a... - 02:48 PM rbd Bug #2457 (Resolved): libvirt: migration fails with rbd in 0.9.11 and 0.9.12
- As reported on libvirt-users:
https://www.redhat.com/archives/libvirt-users/2012-May/msg00088.html
Looks like a... - 11:16 AM Linux kernel client Bug #2395: kernel crash after unmap a rdb device while the cluster is down
- I am unable to reproduce this with the current code. There was a lot of rbd setup/teardown code that got cleaned up ...
- 10:58 AM Linux kernel client Bug #2287: rbd: crashes with 10Gbit network and fio
- Danny Kukawka wrote:
> We used the attached patch to resolve the immediate problem.
That attachment didn't seem ... - 09:07 AM Bug #2446: libceph: corrupt inc osdmap epoch 24630 off 702 (ffff88001e5d876c of ffff88001e5d84ae-...
- Aha, I see the bug. You can apply the following patch and the problem should go away:...
- 02:45 AM Bug #2446: libceph: corrupt inc osdmap epoch 24630 off 702 (ffff88001e5d876c of ffff88001e5d84ae-...
- The monitors deleted older osdmaps from their mondata directories over the weekend, however I managed to reproduce th...
- 04:18 AM Revision d8f4c6d2 (ceph): Merge branch 'stable'
- Conflicts:
src/os/FileStore.cc - 04:12 AM Bug #2267: Ceph client crashed after shutting down one mds and osd
- I fairly often see almost identical crashes. They're triggered by simply restarting an OSD.
Ceph version: 0.46
Ke... - 03:50 AM Revision 62f8f006 (ceph): rbd.xfstests: default to 250mb instead of 100mb
05/20/2012
- 10:34 PM Bug #2454: "rbd info xyz" hanging forever sometimes
- I wrote a little script which calls "rbd ls" and then loops through "rbd info $x" for every volume.
We've about 80 ... - 04:22 PM Bug #2454 (Need More Info): "rbd info xyz" hanging forever sometimes
- How reproducible is this? If you can reproduce with 'debug ms = 20' on the client side, we can be sure it's the osd'...
- 09:18 AM Bug #2454 (Resolved): "rbd info xyz" hanging forever sometimes
- We're running ceph with 3 mon and 21 osds to host about 80 KVM VMs.
Sometimes "rbd info" is hanging forever without ... - 10:33 PM Revision ad663d58 (ceph): v0.47
- 10:33 PM Revision e2e7f583 (ceph): keys: new release key
- New release key for signing packages. Signed by me (the old release key)
so that existing apt keyrings should be suf... - 10:12 PM Revision 5d2ec1d6 (ceph): builder: make reweight helpers static, void
- Signed-off-by: Sage Weil <sage@inktank.com>
- 10:12 PM Revision 63580a90 (ceph): formatter: replace malloc with new
- This will throw exception if allocation failed.
Signed-off-by: Yehuda Sadeh <yehuda@hq.newdream.net> - 10:12 PM Revision 7eb29aee (ceph): check malloc return values
- There were a few places where we didn't check malloc return code.
Signed-off-by: Yehuda Sadeh <yehuda@hq.newdream.net> - 10:12 PM Revision c7daf7a4 (ceph): crush: check malloc results
- Signed-off-by: Yehuda Sadeh <yehuda@hq.newdream.net>
- 09:12 PM Revision 5b39ae7e (ceph): osdmap: set name_pool correctly in build_simple*()
- Backport: dho
Signed-off-by: Sage Weil <sage@inktank.com> - 09:12 PM Revision 033be2b9 (ceph): osdmap: use pg[p]_num accessors
- This lets us ensure that calc_pg_masks() is called when pg[p]_num is set or
modified.
Fixes: #2448
Backport: dho
Sig... - 09:12 PM Revision deccc592 (ceph): filestore: do not test xattrs on fsid file
- This is sufficient to break our fcntl lock, which breaks the check for
other running daemons.
Broken by f03dc34f7e2f... - 09:11 PM Revision bd5fc0a0 (ceph): debian: fix ceph-kdump-copy.init
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:01 PM Bug #2456 (Resolved): librbd: failed LibRBD.TestIOToSnapshot
- ...
- 07:50 PM Revision 31102d31 (ceph): debian: fix ceph-kdump-copy.init
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:09 PM Bug #2446 (Need More Info): libceph: corrupt inc osdmap epoch 24630 off 702 (ffff88001e5d876c of ...
- Is the osdmap/24630 present on all monitors? Is it identical on all of them?
The attachment is 1386 bytes.
The d... - 03:03 PM Bug #2455 (Resolved): debian: lintian errors
- commit:31102d317d7a091f49f9126a6df9087cde0d8118
- 01:07 PM Bug #2455 (Resolved): debian: lintian errors
- W: ceph-kdump-copy: binary-without-manpage usr/bin/ceph-kdump-copy
W: ceph-kdump-copy: init.d-script-missing-start /... - 02:57 PM Feature #2449 (Resolved): dho rsync email
- commit:31102d317d7a091f49f9126a6df9087cde0d8118
- 12:29 PM Linux kernel client Bug #2389: rbd: hung xfstest 67
- again, ubuntu@teuthology:/a/nightly_coverage_2012-05-20-a/2523
05/19/2012
- 11:12 PM Revision 3f38605d (ceph): filestore: test xattrs on temporary file
- There isn't one we can rely on existing yet. Broken by
1314a00798ed4b7ef2f2686f0195c5c53c98c2ce while fixing #2452.
... - 10:08 PM Revision 34e8169e (ceph): monmaptool: fix clitests
- - epoch is always 0
- new options in usage string
Signed-off-by: Sage Weil <sage@inktank.com> - 09:56 PM Revision 1314a007 (ceph): filestore: do not test xattrs on fsid file
- This is sufficient to break our fcntl lock, which breaks the check for
other running daemons.
Broken by f03dc34f7e2f... - 08:24 PM Revision 44fe8c82 (ceph): osdmap: set name_pool correctly in build_simple*()
- Backport: dho
Signed-off-by: Sage Weil <sage@inktank.com> - 08:24 PM Revision ba2488f2 (ceph): osdmap: use pg[p]_num accessors
- This lets us ensure that calc_pg_masks() is called when pg[p]_num is set or
modified.
Fixes: #2448
Backport: dho
Sig... - 06:11 PM Bug #2453 (Resolved): osd/OSD.h: 840: FAILED assert(last_scrub_pg.count(p))
- ...
- 05:06 PM Feature #2405: osd: Make ceph-osd --mkfs idempotent
- 04:44 PM Revision 69e823e5 (ceph): cephtool: restore osdmap handler
- Accidentally removed by 841372494f531c7481cf3eff1b673cedbab4d60d.
Signed-off-by: Sage Weil <sage@inktank.com> - 03:48 PM Bug #2452 (Resolved): filestore: running daemon check broken
- fixed by commit:1314a00798ed4b7ef2f2686f0195c5c53c98c2ce
- 03:48 PM Bug #2452 (Resolved): filestore: running daemon check broken
- the xattr checks in _detect_fs on fsid file break the fcntl lock.
broken by commit:f03dc34f7e2fc1707fa00339b917c0d... - 02:19 PM Bug #2448 (Resolved): osdmap: mapping doesn't work without encoding and decoding
- 02:19 PM Bug #2448: osdmap: mapping doesn't work without encoding and decoding
- fixed by commit:ba2488f238923199534d56a8a86df4e48c2ddd96
- 10:39 AM Bug #2451 (Can't reproduce): qa: networking doesn't always start after reboot
- ...
- 12:26 AM Revision 3f637b81 (ceph): mon: show quorum names in health warning
- Names, not just ranks (which nobody cares about).
Signed-off-by: Sage Weil <sage@inktank.com> - 12:26 AM Revision 360930a0 (ceph): qa: use full paths for mon bootstrap tests
- Signed-off-by: Sage Weil <sage@inktank.com>
05/18/2012
- 11:23 PM Revision ece33e5b (ceph): monmaptool: add --generate, --filter-initial options
- Generate a monmap from conf/args, and apply the mon initial hosts to it.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:23 PM Revision 359fc7bb (ceph): admin_socket: match first word against command
- This lets us do commands that take arguments. We assume space is the
separator.
Signed-off-by: Sage Weil <sage@inkt... - 11:23 PM Revision aa1ee9be (ceph): cephtool: multiword admin socket commands
- Slurp up anything on the command line that isn't an option, so that you can
do
ceph --admin-socket /path/to/asok fo... - 11:23 PM Revision 824122d7 (ceph): mon: all starting with empty monmap
- It's okay to start up with an empty monmap. It just means that you must:
* set 'mon initial members'
* feed hints... - 11:23 PM Revision b5bec554 (ceph): mon: 'add_bootstrap_peer_hint <addr>' via admin socket
- Let user add peer hints via the admin socket while bootstrapping. Once we
have joined a cluster this command is igno... - 11:23 PM Revision fc447bd4 (ceph): mon: add peers probing us to extra peer list
- If we are probed by another monitor, add them to our extra probe list. This
lets us rely on the active probe/reply to... - 11:23 PM Revision a5eba525 (ceph): mon: set our addr when populating monmap with initial members
- If the seed monmap doesn't contain us and we are populating it with
initial members, and one of those members is us, ... - 11:23 PM Revision dfac9250 (ceph): mon: learn peer addrs when probing
- If we have a monmap with dummy addrs for initial members, and we probe one
of those peers and learn their address, up... - 11:23 PM Revision 6551298f (ceph): mon: simplify/clean up dummy addrs used for initial members
- Use a complete blank IP, but set the nonce. This way
entity_addr_t::is_blank_ip() works. We're also outside of the ... - 11:23 PM Revision 434c158f (ceph): mon: send join message if we are in monmap with blank addr
- Being in the monmap with a blank ip is possible if we were a initial member
seed but weren't part of the first electi... - 11:23 PM Revision 3920ce30 (ceph): mon: add -or- update blank mon addr with join command
- Update blank addrs with the join command (in addition to adding new mons).
Signed-off-by: Sage Weil <sage@inktank.com> - 11:23 PM Revision 777b739b (ceph): qa: add mon test for seeding bootstrap mon peers via admin socket
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:23 PM Revision 3c771924 (ceph): mon: always mark seed/mkfs monmap as epoch 0
- This isn't strictly necessary, since the MonmapMonitor resets this on first
commit, but it's cleaner.
Signed-off-by:... - 11:23 PM Revision a68c224a (ceph): mon: move initial_members filtering into MonMap method
- This is cleaner. Also, we can perform/test this functionality from the
monmaptool.
Signed-off-by: Sage Weil <sage@i... - 11:23 PM Revision 1473ef43 (ceph): monmap: move build_initial() from MonClient
- This belongs here, not in MonClient.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:23 PM Revision dc9122cb (ceph): mon: limit initial quorum to mon_initial_members
- This is a two-stage process.
* If we start up, and have never joined a quorum, and initial members are
specified... - 11:23 PM Revision 6c8214fb (ceph): mon: include quorum member names in 'mon stat'
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:23 PM Revision cba3ca7b (ceph): mon: ignore election messages from outside monmap
- These shouldn't(tm) happen with new code, but with old code they do. And
if we get them, elector can try to monmap->... - 11:23 PM Revision 0b74b8f6 (ceph): qa: add a bunch of mon bootstrap tests
- These's are comprehensive because a lot of the startup logic is about
picking a local address, and it's difficult to ... - 11:23 PM Revision 59ccf9ce (ceph): mon: clean up initial_members code slightly
- Don't recalc rank unless we need to.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:23 PM Revision b67bc0bc (ceph): mon: drop mon_initial_hosts
- Bad idea.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:23 PM Revision 262d5e2f (ceph): monmap: move monmap generation from ip/host list into MonMap
- Also generalize the prefix so it's explicit in the caller and a bit less
magic.
Signed-off-by: Sage Weil <sage@inkta... - 11:23 PM Revision c52a4f72 (ceph): monmaptool: don't increment epoch on modification
- This just confuses things, because a manually manipulated map might have
some epoch number that bears no relation to ... - 11:23 PM Revision 0088699f (ceph): mon: take probed peer's monmap if it has ever joined a quorum
- If we probe a peer and their monmap has actually been part of a started
cluster/quorum, and ours hasn't, take theirs.... - 11:23 PM Revision c7ba9fea (ceph): mon: use current monmap for initial quorum
- This makes a bit more sense. Don't use the seed monmap, but use the one
we ended up with when we formed our first qu... - 11:23 PM Revision e1d34cc4 (ceph): mon: add set of additional peers to probe
- Probe an additional set of peers, not included in the monmap.
Incomplete; we'll still assume they *can* form a quoru... - 11:23 PM Revision 585adcf1 (ceph): mon: only include monmap mon's in outside_quorum, new quorum
- Only include mons that are in the monmap in the outside_quorum set, which
we use to determine if we can form a new qu... - 11:23 PM Revision e52093c5 (ceph): mon: clean up "joined" flag
- - check flag on init, keep in memory
- set flag only when we are active and have a committed monmap. i.e., when
we... - 11:23 PM Revision 5b186657 (ceph): mon: pass has_ever_joined in probe messages
- This will allow monitors to behave appropriately with that important info..
Signed-off-by: Sage Weil <sage@inktank.com> - 11:21 PM Revision 7e6be867 (ceph): mon: discover peer names during probe
- Fill in noname- entires in initial generate monmap (e.g., from mon_host)
with real peer names as we discover them.
T... - 11:21 PM Revision e197e79d (ceph): mon: add initial quorum config options
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:21 PM Revision 52c28456 (ceph): monclient: name resolved mon hosts noname-* too
- This would allow you to do
mon host = foo, bar, baz
or
mon host = multiarecordhostname.foo.com
and have the g... - 10:07 PM Revision 2b5ee3e0 (ceph): mon: include health in status
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:08 PM Revision 7d6facc0 (ceph): mon: move health check into helper function
- Signed-off-by: Sage Weil <sage@inktank.com>
- 08:54 PM Revision 09978a8a (ceph): Merge remote-tracking branch 'gh/wip-mon-refactor'
- Reviewed-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 07:07 PM Revision d599eded (ceph): mon: fix a bug that was preventing us from subscribing to some log levels.
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 06:32 PM Revision 20df4324 (ceph): mon: fix the priority filter.
- When a given log level L was specified, we would reply with all the
messages of "level L and below"; for instance, fo... - 05:59 PM Revision 9fb3cd28 (ceph): ceph-authtool: fix clitests
- No more --bin option.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:59 PM Revision 4832c46e (ceph): ceph-authtool: fix usage
- No more -b, --bin.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:55 PM Revision ca987d38 (ceph): msgr: If the Connection has a bad Pipe*, replace it with a good one.
- This is a little weak, since it doesn't fill in the Connection's data
when we initially create the new Pipe. But doin... - 05:55 PM Feature #2450 (Resolved): dho git commit emails
- Send git commit emails for dho branch to dho folk. Just need some git hook to send this to the dho list for the dho ...
- 05:54 PM Feature #2449 (Resolved): dho rsync email
- 05:52 PM Bug #2420 (Resolved): ceph crash while under iogen load
- yay!
- 04:49 PM Bug #2420: ceph crash while under iogen load
- Updated the cluster with the kernel 3.3.0-ceph-00110-g1d4a9bf and ran iogen for 2 hrs, cleared the mounted file store...
- 05:35 PM Linux kernel client Bug #2447 (Resolved): prepare_write_connect NULL pointer dereference
- this was a bug in the new messenger refactor, pushed an updated commit:3da54776e2c0385c32d143fd497a7f40a88e29dd
- 05:00 PM Linux kernel client Bug #2447 (Resolved): prepare_write_connect NULL pointer dereference
- ...
- 05:09 PM Bug #2448 (Resolved): osdmap: mapping doesn't work without encoding and decoding
- adamcrume on irc was having trouble with this sample program only mapping to the same two osds for all objects. encod...
- 04:23 PM Feature #2418 (Fix Under Review): mon: Take mandatory initial quorum members from ceph.conf
- 04:23 PM Feature #2419 (Fix Under Review): mon: Take peer hints via admin socket
- 04:15 PM Feature #2423 (Resolved): gceph: remove it
- 04:15 PM Feature #717 (Resolved): cephtool: make -s/-w use subscribe instead of paxos watch; deprecate pax...
- 12:54 PM Feature #717: cephtool: make -s/-w use subscribe instead of paxos watch; deprecate paxos watch
- Sage Weil wrote:
> The --watch-error etc. options don't seem to work.. otherwise this is basically ready. Pushed a ... - 08:38 AM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- Hi Oliver,
The fix will be in 0.47, which should be out in the next couple days. Glad to hear it's fixed!
Josh - 06:00 AM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- Hi Josh,
yeah, seems to be fixed. When do we have to expect it in a more "stable" release?
Thnx for all efforts... - 03:29 AM Revision b180ac25 (ceph): Merge branch 'wip-mon-keyring'
- Reviewed-by: Greg Farnum <greg@inktank.com>
- 03:23 AM Revision 9a1d757b (ceph): mon: only log map state changes from primary
- We still get noise after elections, but only from the winner.
Signed-off-by: Sage Weil <sage@inktank.com> - 03:23 AM Revision 9125e908 (ceph): mon: monmap first from 'status' command
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:23 AM Revision f943f34b (ceph): debian: don't package gceph
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:23 AM Revision a586bdf0 (ceph): ceph.spec: don't package gceph
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:23 AM Revision 89316566 (ceph): remove gceph
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:23 AM Revision 6b76ac3a (ceph): Makefile: drop MMonObserve*
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:23 AM Revision 13a9a152 (ceph): mon: aesthetic changes to map change log entries
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:23 AM Revision 78d0d097 (ceph): cephtool: aesthetic change to ceph -w output
- * no seq
* name, not ip addr
Signed-off-by: Sage Weil <sage@inktank.com> - 02:14 AM Bug #2446 (Resolved): libceph: corrupt inc osdmap epoch 24630 off 702 (ffff88001e5d876c of ffff88...
- Ceph version: 0.46
Kernel client version: Debian's linux-image-3.3.0-trunk-amd64=3.3.4-1~experimental.1 patched so t...
05/17/2012
- 11:15 PM Revision 84137249 (ceph): cephtool: Slight rework and pub/sub log support.
- We reworked the code a bit to accommodate the introduction for the log
monitor's publish/subscribe mechanisms. With t... - 11:15 PM Revision 035bb125 (ceph): src: get rid of the Observers throughout the code base.
- This is a big patch that will remove all references to the observers
throughout the code, including a complete remova... - 10:55 PM Revision 6fe96cb9 (ceph): mon: Add publish/subscribe capabilities to the log monitor and status cmd.
- This patch allows us to stir away from the monitor's observer mechanism,
by using instead the already existing publis... - 09:24 PM Bug #2441 (Resolved): haproxy, rgw: returns 502
- Ok, hasn't reproduced anymore. Resolving.
- 04:37 PM Bug #2441: haproxy, rgw: returns 502
- Our haproxy configuration was broken (values are in milliseconds, not in seconds):...
- 09:37 AM Bug #2441 (Resolved): haproxy, rgw: returns 502
- We see occasional 502 responses that originate at haproxy. The apache and rgw logs don't have any indications about s...
- 09:22 PM Feature #2327 (Resolved): mon: use external keyring for inter-mon auth
- 09:19 PM Feature #2405 (In Progress): osd: Make ceph-osd --mkfs idempotent
- 09:18 PM Feature #717: cephtool: make -s/-w use subscribe instead of paxos watch; deprecate paxos watch
- The --watch-error etc. options don't seem to work.. otherwise this is basically ready. Pushed a bunch of cleanups to...
- 08:56 PM Revision 572c7560 (ceph): doc: fix sizes in librbd python examples
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 08:06 PM Revision a5dd3872 (ceph): Merge remote branch 'origin/wip-misc'
- Reviewed-by: Sage Weil <sage@inktank.com>
- 07:57 PM Revision e7212714 (ceph): doc: fix link to hardware recommendations section
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 07:57 PM Revision 25e7889a (ceph): doc: fix formatting and a couple config defaults
- ``ceph.conf`` needed spaces around it to be parsed correctly.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 06:20 PM Feature #2419: mon: Take peer hints via admin socket
- wip-quorum
passing my test, qa/mon/bootstrap/initial_members_asok.sh - 04:54 PM Revision 62db60be (ceph): FileJournal: Clarify log message on journal too small.
- The OSDs tolerate a full journal, it will just result in a latency
spike. Make the error message say operation will c... - 03:39 PM Feature #2418: mon: Take mandatory initial quorum members from ceph.conf
- dropped mon_initial_hosts.. that was a bad idea.
see wip-quorum - 03:35 PM CephFS Bug #2445 (Can't reproduce): crash when removing a non-empty directory
- From http://permalink.gmane.org/gmane.comp.file-systems.ceph.devel/6237:...
- 03:11 PM CephFS Bug #2444 (Can't reproduce): null pointer deference in ceph_d_prune inside kvm
- From http://permalink.gmane.org/gmane.comp.file-systems.ceph.devel/6180:...
- 10:09 AM Bug #2443 (Resolved): Anyone can list all keys, even with caps mon 'allow rwx' and not 'allow *'
- Caps are kind of pointless if I can just ask for any secret I want.
ubuntu@inst03:~$ sudo ceph --name=osd.4 --keyr... - 10:06 AM Bug #2442 (Resolved): bash completion is broken
- "ceph --keyring=/et" TAB TAB TAB nothing happens. Even the built-in, non-customized completion for "doesnotexist --ke...
- 09:17 AM Feature #2440 (Resolved): osd: understand btrfs performance
- 08:46 AM Bug #2420: ceph crash while under iogen load
- it should still be reproducible. I left the configuration up and I was able to reproduce it a couple of days ago.
- 05:24 AM Linux kernel client Bug #2439 (Resolved): ceph-client: auth: handle null verify_authorizer_reply method
- I added code to the client messenger avoid dereferencing a null
auth_client->ops or auth_client->ops->(method) point... - 05:18 AM Linux kernel client Cleanup #2438 (Closed): ceph-client: use BUG_ON() for null auth_client->ops pointers
- I added code to in the client messenger to verify auth_client->ops
and auth_client->ops->(method) were non-null befo... - 01:37 AM Revision 96426fa4 (ceph): ReplicatedPG: fill in data_off in OSDOpReply
- The assignment was accidentally made unreachable by
ff55d2f310312bb5390326dcc35961d39ccad416
Signed-off-by: Josh Dur... - 12:43 AM Revision 04a75d29 (ceph): Merge branch 'next'
- 12:42 AM Revision 4277d4d3 (ceph): Merge branch 'wip-oc-error-handling' into next
- Reviewed-by: Sage Weil <sage@inktank.com>
- 12:02 AM Revision 1ec3245d (ceph): config: improve comments for rbd caching options
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 12:02 AM Revision 49c6ad44 (ceph): librbd: check for cache flush errors
- Return errors from flushing to the caller. Warn
if an error occurs during invalidation, but don't retry,
since the hi...
05/16/2012
- 11:55 PM Revision ef71d85e (ceph): Merge remote-tracking branch 'gh/wip-mon-encoding'
- Reviewed-by: Greg Farnum <greg@inktank.com>
- 11:53 PM Revision f161c623 (ceph): Merge branch 'wip-upstart'
- Reviewed-by: Tommi Virtanen <tv@inktank.com>
- 11:37 PM Revision 5e3bc57e (ceph): init-radosgw: skip sections without host= line
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:36 PM Revision 6d4e79ac (ceph): mon: make 'auth get-key <name>' return key
- Same as 'print-key'.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:36 PM Revision ec732816 (ceph): keyring: always encode in plaintext
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:36 PM Revision eaea7aa9 (ceph): keyring: drop binary encoder
- Drop the keyring encode method, and binary encoder.
Don't just encode in plaintext because we assume we get the whol... - 11:36 PM Revision 28863584 (ceph): init-ceph: ignore sections without 'host' defined
- This will make it easier for sysvinit and upstart to coexist.
We will break existing users who have a separate .conf... - 11:36 PM Revision d0790186 (ceph): mon: make 'auth get <name>' dump plaintext keyring
- This is friendlier.
Fixes: #2406
Signed-off-by: Sage Weil <sage@newdream.net> - 11:36 PM Revision 66e27375 (ceph): auth: EntityName ==, != operators
- Signed-off-by: Sage Weil <sage@newdream.net>
- 11:36 PM Revision c9b220ff (ceph): mon: 'auth get-or-create-key name capsys capval capsys2 capval2 ...'
- Fixes: #2406
Signed-off-by: Sage Weil <sage@newdream.net> - 11:36 PM Revision 86f24868 (ceph): mon: 'auth list' is a read-only operation
- Signed-off-by: Sage Weil <sage@newdream.net>
- 11:32 PM Revision 23deb43b (ceph): ObjectCacher: handle write errors
- If a write error occurs, mark the BufferHead dirty again, and
pass the return value to the completion. This makes flu... - 11:31 PM Revision 59c68165 (ceph): ObjectCacher: propagate read errors to the caller
- Previously the return value of a read operation was ignored. Now a
read error sets the error field, and changes the ... - 11:10 PM Revision 24fddf21 (ceph): ceph-dencoder: add AuthMonitor::Incremental
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:06 PM Revision 7f148545 (ceph): mon: use new encoder macros for AuthMonitor::Incremantal
- Signed-off-by: Sage Weil <sage@inktank.com>
- 11:04 PM Revision 7aaccacf (ceph): features: explicitly use CEPH_FEATURES_ALL instead of -1
- Make it clear to readers what that -1 is all about. Alex would be proud.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:37 PM Revision 51564955 (ceph): mon: fix mon removal check
- Only take our absence from the monmap to mean that we were removed if we
were ever a member in the first places.
Thi... - 10:16 PM Revision 0f7bf5fd (ceph): rbd: allow importing from stdin
- This patch allows importing images from stdin with the following command:
rbd import --size=<size in MB> - [dest-... - 09:47 PM Feature #2426: precise packages for apache2, fastcgi
- Pretty sure i built this on pudgy, whose disk has just died.
I'd apt-get source these and verify we can rebuild ... - 07:14 AM Feature #2426: precise packages for apache2, fastcgi
- Ok, I just rsynced the dho packages (manually built) over to the gitbuilder url. At the very least, need to document...
- 07:14 AM Feature #2426: precise packages for apache2, fastcgi
- Either set up a gitbuilder, or build the packages manually and put in a repo, or just use the dho packages+repo (buil...
- 07:10 AM Feature #2426 (Resolved): precise packages for apache2, fastcgi
- 09:44 PM Bug #2420 (Fix Under Review): ceph crash while under iogen load
- I think fix-unregister-race will fix this.. Alex or Yehuda, does that make sense?
Hopefully the crash is reproduci... - 12:03 PM Bug #2420: ceph crash while under iogen load
- The iogen command used was:
sudo iogen -s 2g -b 128k -t 1 -d /mnt/osd -n 5 - 08:50 PM Feature #2295: make qemu cache=writeback,writethrough option turn on librbd caching
- Looks good! Reviewed-by: etc.
- 03:43 PM Feature #2295 (Fix Under Review): make qemu cache=writeback,writethrough option turn on librbd ca...
- See wip-cache
- 08:46 PM Bug #2437 (Resolved): osd: very slow during recovery
- On congress, s3tests runtimes go from 1-2 minutes to 7+ minutes when there is any recovery going on. This appears to...
- 08:43 PM Revision 48d97fe7 (ceph): doc: add ioctx parameter to rbd pybind examples.
- This parameter is required and was missing. And fix a typo
- 08:40 PM Revision 1109c132 (ceph): librados: avoid overflow in the return value of reads
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 08:40 PM Revision 2f1f8088 (ceph): ObjectCacher: only perfcount reads requested by the client
- _readx is called again after each bh is read by C_RetryRead. This
resulted in the read being counted many times for t... - 08:15 PM Revision 4bc63ced (ceph): msgr: look for a pre-existing Pipe if the Connection doesn't have one.
- We may have already established a new Connection to the peer in question.
Signed-off-by: Greg Farnum <gregory.farnum... - 08:07 PM Revision 1c53f34b (ceph): osd: include age in 'slow request' messages
- Make log messages easier to interpret, without subtracting timestamps in
your head.
Signed-off-by: Sage Weil <sage@i... - 07:55 PM Revision 2b446620 (ceph): workloadgen: Add transaction throughput infos.
- Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
- 06:08 PM Feature #2290 (Resolved): ObjectCacher: handle read/write errors
- 02:01 PM Feature #2290 (Fix Under Review): ObjectCacher: handle read/write errors
- Updated wip-oc-error-handling to use a separate BufferHead state for read errors, and just reset the state to dirty o...
- 04:53 PM Cleanup #2344 (Resolved): convert Monitor maps to use ENCODE_START
- 04:50 PM Cleanup #2344: convert Monitor maps to use ENCODE_START
- Looks good. Reviewed-by: Greg Farnum. :)
- 04:10 PM Cleanup #2344 (Fix Under Review): convert Monitor maps to use ENCODE_START
- 02:23 PM Cleanup #2344: convert Monitor maps to use ENCODE_START
- MonCaps look good; I'm a bit concerned about the "all features" default used for the MonMap (left a note on GitHub) b...
- 04:52 PM Feature #2407 (Resolved): auth: "ceph auth get NAME"
- 04:52 PM Feature #2404 (Resolved): init-ceph: Make /etc/init.d/ceph ignore entries without explicit host= ...
- 04:52 PM Feature #2406 (Resolved): auth: "ceph auth get-or-create NAME CAPS.."
- 03:36 PM Bug #2436 (Resolved): mon: suicides when trying to join an existing quorum
- commit:515649558d5edebfd705e63bc34cd74d2db1f682
- 03:18 PM Bug #2436: mon: suicides when trying to join an existing quorum
- This is because when you provide a new monitor with a monmap that includes both itself and existing monitors, the new...
- 03:11 PM Bug #2436 (Resolved): mon: suicides when trying to join an existing quorum
- I have mon.single running happily at 192.168.122.91. On 192.168.122.159, I run (using the right values for key and fs...
- 02:39 PM Feature #717 (In Progress): cephtool: make -s/-w use subscribe instead of paxos watch; deprecate ...
- 01:54 PM Documentation #2388 (Resolved): librbd python doc lacks ioctx parameter to rbd.Image() constructor
- commit:48d97fe79634d4d4bb7ea2237083c3cd694ff3fe
- 01:51 PM Cleanup #2435 (Resolved): Remove binary keyring support
- Plaintext keyrings have been supported (and default) for almost a year and a half (#705, cfae10b8f8b0d91f37dc6eb72f3b...
- 01:32 PM rgw Bug #2433 (Resolved): rgw: failing atomic reads/writes
- Problem was in the s3-tests test suite.
Fixed, commit:adabd0ba7def8fc12e00b2c19a37d5936d53eff6 and commit:f1f86a0d... - 11:39 AM rgw Bug #2433: rgw: failing atomic reads/writes
- Triggered on boto 2.4.0.
- 11:18 AM rgw Bug #2433 (Resolved): rgw: failing atomic reads/writes
- ...
- 01:04 PM Linux kernel client Bug #2392: First read of symlink after ceph filesystem mounted gives error
- The problem is the lookup open intents stuff. We try to do a lookup + open, but it ends up that the lookup result is...
- 11:45 AM Linux kernel client Bug #2392: First read of symlink after ceph filesystem mounted gives error
- Okay, this looks to me like it has to be a problem with the kernel client. The MDS definitely knows it's a symlink at...
- 11:37 AM rgw Bug #2434 (Duplicate): rgw: failing readwrite test
- Duplicates #2433.
- 11:18 AM rgw Bug #2434 (Duplicate): rgw: failing readwrite test
- ...
- 10:36 AM Linux kernel client Cleanup #2432 (Resolved): ceph-client: messenger: refactor to simplify state model
- There is a mix of states and flags used in the client messenger code
to track what's going on. The result is a litt... - 10:21 AM Feature #2431 (Duplicate): teuthology: qemu + rbd testing
- 10:13 AM Feature #1711: chef: multiple monitor support
- Depends: #2406
- 10:08 AM Feature #2241 (Rejected): upstart
- 10:05 AM Feature #2399 (In Progress): qa: haproxy + rgw + jenkins
- 09:55 AM Subtask #2430 (Resolved): simplify pg removal
- PG resurrection complicates the implementation and can be removed with little harm.
- 09:03 AM Feature #2428: auth: revise auth config params
- We need to keep backwards compatibility with auth supported so libvirt doesn't break.
- 08:03 AM Feature #2428 (Resolved): auth: revise auth config params
- new:
* auth cluster required = [cephx] what mon requires of mon, mds, osd daemons
* auth service required = [cephx]... - 08:58 AM Linux kernel client Bug #2429: ceph-client: verify_authrizer_reply con method never called
- Whoops, forgot to fill in a real subject/title. Not sure how to
fix it either. - 08:54 AM Linux kernel client Bug #2429 (Resolved): ceph-client: verify_authrizer_reply con method never called
- Both ceph_connection_operations and ceph_auth_client_ops define
a verify_authorizer_reply method.
The only caller... - 07:11 AM Feature #2427: precise gitbuilder http
- * thttpd dne on precise. make a package, or use apache, nginx, or something.
- 07:10 AM Feature #2427 (Resolved): precise gitbuilder http
- 06:22 AM Linux kernel client Bug #2424 (Resolved): ceph-client: messenger: badness in prepare_write_connect()
- At the end of prepare_write_connect() there is a call to prepare_connect_authorizer().
That function gets an authori...
05/15/2012
- 09:00 PM Feature #2423: gceph: remove it
- This is the only remaining user of paxos observer framework. We could easily add a "pgmap" subscribe to support it, ...
- 08:59 PM Feature #2423 (Resolved): gceph: remove it
- 08:56 PM Feature #1651 (Resolved): command line tool to interact with admin socket
- ceph --admin-daemon <socket> <command ...>
- 08:05 PM Feature #1488 (Resolved): chef: spec/break down osd addition, replacement
- 08:02 PM RADOS Feature #2422 (Resolved): crush: test that mapping result is uncorrelated
- Verify that crush outputs are not correlated between devices. e.g., that all items with primaries on device A also g...
- 07:57 PM RADOS Feature #2421 (Resolved): crush: quantitatively validate mapping quality
- - measure variance of resulting mapping vs expected (based on weights)
- give some quantitative comparison with expe... - 06:16 PM Revision 3d2d0ecd (ceph): Additional documentation theme changes:
- 1) Adjust h2 tags so that section titles are visually differentiated
2) Add 1.5em of margin to all pre blocks and tab... - 05:07 PM Bug #2420 (Resolved): ceph crash while under iogen load
- ken.franklin@inktank.com was running a cluster with three nodes, one OSD and one monitor per node, and one mds.
ceph... - 04:26 PM Linux kernel client Bug #2392: First read of symlink after ceph filesystem mounted gives error
- Sorry - remembering to enable debugging that would have been more helpful! Logs with debugging turned on attached.
- 09:51 AM Linux kernel client Bug #2392: First read of symlink after ceph filesystem mounted gives error
- Mark, can you repeat these with debug logging turned up? It'll take a fair bit of disk space but there's not very muc...
- 11:27 AM Feature #1711: chef: multiple monitor support
- Depends: #2418 #2419
Current architecture:
Assumption: mon id is always chef nodename (= short hostname)
Assum... - 11:19 AM Feature #2419 (Resolved): mon: Take peer hints via admin socket
- Allow starting a mon with an empty monmap and "mon hosts". Make it
take "peer hints" via the admin socket. Peer hint... - 11:16 AM Feature #2418 (Resolved): mon: Take mandatory initial quorum members from ceph.conf
- Make it so that forming an initial quorum is only possible if the
majority of members from a "initial members" are p... - 11:04 AM Bug #2316 (Resolved): rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g30...
- More than 100 runs with no problems, I'd say this is fixed. commit:d7343814a01257a4f727fdfc752361b930ab5719
- 09:37 AM devops Feature #2417: chef: support radosgw
- Depends on #2415.
- 09:36 AM devops Feature #2417 (Rejected): chef: support radosgw
- 09:36 AM devops Feature #2416: chef: support mds
- Depends on #2414.
- 09:36 AM devops Feature #2416 (Rejected): chef: support mds
- 09:36 AM devops Feature #2415 (Resolved): upstart: support radosgw
- 09:35 AM Feature #2414 (Resolved): upstart: support mds
- 09:08 AM Feature #2413 (Resolved): qa: Test co-existence of sysvinit and upstart, 3: upstart controlled
- Depends on #2404. Depends on branch chef-2. TODO merge not available currently.
Deploy using ceph-cookbooks.git.
En... - 09:07 AM Feature #2412 (Resolved): qa: Test co-existence of sysvinit and upstart, 2: sysvinit controlled, ...
- Depends on #2404. Depends on branch chef-2. TODO merge not available currently.
Ensure osd data disks do NOT use GPT... - 09:07 AM Feature #2411 (Resolved): qa: Test co-existence of sysvinit and upstart, 1: sysvinit controlled, ...
- Depends on #2404. Depends on branch chef-2. TODO merge not available currently.
Ensure osd data disks do NOT use GPT... - 04:15 AM Revision 067cd5fb (ceph): monmap: new encoder
- Introduce a feature bit so that we can interoperate with older
peers.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:03 AM Revision 31cb3aa6 (ceph): mon: conditonally encode monmap for peer mon during probe
- This might be sufficient to let monitors with different versions of the
monmap encoding interoperate, but I'm too laz... - 04:01 AM Revision a9f5d774 (ceph): monmap: use feature bits and single encode() method
- Instead of selecting an encode method in the caller, use a normal features
argument to encode() and branch there.
Le... - 03:53 AM Revision 31dc3945 (ceph): mon: new encoding for MonCaps
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:39 AM Bug #2379: Mon crash after start
- The problem occurred again....
- 03:30 AM Revision 5b0c7043 (ceph): Merge branch 'stable'
- Conflicts:
src/osdc/Objecter.cc - 03:23 AM Revision d7343814 (ceph): Objecter: don't throttle resent linger ops
- Throttling is intended to stop the caller from submitting too many
requests, not blocking requests that are being res... - 03:22 AM Revision ac85b9e8 (ceph): objecter: ping connections with lingering ops every tick()
- Also, make sure that we initialize linger_op.session once
we have it.
Signed-off-by: Yehuda Sadeh <yehuda@hq.newdrea... - 03:13 AM Revision 7830f859 (ceph): mon: use external keyring for mon->mon auth
- - Feed our keyring into the auth methods.
- Do not fail to build a ticket for type MON when we don't have a cap; it
... - 03:13 AM Revision 7be78101 (ceph): mon: keep mon. secret in an external keyring
- - Keep the mon. key in a separate keyring files, "keyring", in the mon
data dir.
- During init, if we don't find th...
05/14/2012
- 09:34 PM rbd Bug #2410 (Closed): hung xfstest #68
- ...
- 09:22 PM Feature #2408: librbd: track latency with perfcounters
- 04:49 PM Feature #2408 (Resolved): librbd: track latency with perfcounters
- We're currently counting read/write/etc op and byte counts. We also want to know latencies in each category.
(For... - 09:07 PM Bug #2409 (Resolved): osd: pgs stuck in active
- congress has several pgs stuck in active, and objects counted as degraded. looks like we missed some corner case in ...
- 07:57 PM Revision 784d1d3a (ceph): Documentation theme updates:
- 1) Background color in navbar
2) Correct body fonts
3) Document pane background (below the content) - 06:34 PM Revision 84f335a6 (ceph): Merge branch 'wip-osdmap'
- Reviewed-by: Greg Farnum <greg@inktank.com>
- 06:27 PM Revision ab717bc5 (ceph): Updated link to publications
- 06:24 PM Revision 5922e2c2 (ceph): crush: pass weight vector size to map function
- Pass the size of the weight vector into crush_do_rule() to ensure that we
don't access values past the end. This can... - 06:24 PM Revision 376f0d50 (ceph): crush: adjust max_devices appropriately in insert_item()
- If we insert a new item, make sure max_devices is still the max id + 1.
Signed-off-by: Sage Weil <sage@inktank.com> - 06:24 PM Revision 320d1ebf (ceph): mon: fail 'osd crush set ...' is osd doesn't exist
- If an osd doesn't exist, don't let users add/update it in the crush map.
Signed-off-by: Sage Weil <sage@inktank.com> - 04:47 PM Linux kernel client Bug #2392: First read of symlink after ceph filesystem mounted gives error
- Okay, no guarantees but I will try and check this out at least briefly in the next day or two. :)
- 04:37 PM Linux kernel client Bug #2392: First read of symlink after ceph filesystem mounted gives error
- Ah - good point, no I had not updated max_mds. I redid the setup with 1 mds and 1 osd. Same issue, logs attached.
- 02:26 PM Linux kernel client Bug #2392: First read of symlink after ceph filesystem mounted gives error
- I notice looking at your conf file that you have 3 MDSes. Are they all active? (ie, did you increase max_mds to 3)
I... - 04:31 PM Feature #2327: mon: use external keyring for inter-mon auth
- wip-mon-keyring
- 03:44 PM Feature #2327: mon: use external keyring for inter-mon auth
- Which branch?
- 03:23 PM Feature #2327 (Fix Under Review): mon: use external keyring for inter-mon auth
- 03:28 PM Bug #2393 (Duplicate): objecter: dropping messages (old connection being used)
- 02:34 PM Bug #2393: objecter: dropping messages (old connection being used)
- Just to make sure we're on the same page: there is nothing in that snippet indicating that there is an active Connect...
- 03:27 PM RADOS Bug #2401 (Resolved): Remove OSD from CRUSH map fails if OSD Running
- this bug was just fixed by wip-osdmap branch, yay!
- 10:28 AM RADOS Bug #2401 (Resolved): Remove OSD from CRUSH map fails if OSD Running
- Removing an OSD from the CRUSH map requires the user to stop the OSD first. If the user doesn't stop the OSD and runs...
- 03:23 PM Cleanup #2344 (In Progress): convert Monitor maps to use ENCODE_START
- 11:23 AM Cleanup #2344: convert Monitor maps to use ENCODE_START
- The PGMap changes there are fine, but there are several others:
AuthMonitor::Incremental
MonCaps
MonMap
OSDMap
... - 03:20 PM Feature #2404 (Fix Under Review): init-ceph: Make /etc/init.d/ceph ignore entries without explici...
- wip-upstart
- 02:09 PM Feature #2404 (Resolved): init-ceph: Make /etc/init.d/ceph ignore entries without explicit host= ...
- 03:20 PM Feature #2407 (Fix Under Review): auth: "ceph auth get NAME"
- wip-upstart
- 02:12 PM Feature #2407 (Resolved): auth: "ceph auth get NAME"
- Comparable to "ceph auth list" and filtering the results, but doesn't leak as many secrets.
Not a dependency of anyt... - 03:19 PM Feature #2406 (Fix Under Review): auth: "ceph auth get-or-create NAME CAPS.."
- wip-upstart
- 02:11 PM Feature #2406 (Resolved): auth: "ceph auth get-or-create NAME CAPS.."
- If the key does not exist, atomically generate a key and add it to the monitor. Output the new key.
If the key exist... - 03:18 PM Feature #2290 (In Progress): ObjectCacher: handle read/write errors
- Sage actually reviewed this a while ago, and I was changing it to separate the read/write errors and not return the w...
- 02:46 PM Feature #2290: ObjectCacher: handle read/write errors
- This looks okay to me, although I didn't check it for comprehensive-ness (just looked at the diff).
- 02:20 PM CephFS Bug #2385: max mds = 2, mds hang and crash
- The full Ceph filesystem is not currently well-tested, but if you can recreate this with MDS logging on and post the ...
- 01:09 PM CephFS Bug #2385: max mds = 2, mds hang and crash
- ceph hate symlink. :)
i can't rsync debian and ubuntu mirror from local disk to cephfs.
allways, allways when... - 02:10 PM Feature #2405 (Resolved): osd: Make ceph-osd --mkfs idempotent
- This will help upgrades from versions where mkfs did not set the "ready" flag. Without this, the upgrade would re-mkf...
- 01:31 PM Bug #2396 (Resolved): don't fail on mismatched size CRUSH and OSD maps
- Merged to master in commit:84f335a6894db6d6b993d9cba584b4c3ab5365d0
- 01:31 PM Bug #2397 (Resolved): mon: prevent addition of CRUSH items past the max_osd
- Merged to master in commit:84f335a6894db6d6b993d9cba584b4c3ab5365d0
- 01:10 PM Subtask #2403 (Resolved): remove osd pointer from PG
- In order to more clearly delineate the osd methods which can be correctly called from PG, those services should be pr...
- 01:09 PM Subtask #2402 (Resolved): audit calls into osd from pg for locking correctness
- With pg peering in a workqueue, we can't hold the osd_lock while handling peering events. Thus, pg calls into the os...
- 12:38 PM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- Er, the important hung thread:...
- 12:28 PM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- The deadlock was due to throttling of resent linger requests during map changes. wip-objecter-throttle should fix thi...
- 11:29 AM Bug #2379: Mon crash after start
- Sage, did you check out the logging for slurping yet?
Maciej, do you still need instructions on core files? - 10:11 AM Bug #2022 (Need More Info): osd: misdirectect request
- not fixed after all, failed again on
ubuntu@teuthology:/a/nightly_coverage_2012-05-13-b/1268 - 10:08 AM rbd Bug #2400 (Resolved): xfstest: failed #84
- 2012-05-13T01:51:49.654 INFO:teuthology.orchestra.run.out:084 - output mismatch (see 084.out.bad)
2012-05-13T01:5... - 09:47 AM Feature #2399 (Resolved): qa: haproxy + rgw + jenkins
- 09:45 AM devops Feature #2398 (Rejected): chef: external osd journal support
- 09:35 AM RADOS Feature #2268 (Resolved): crush: update item's position in crush map
- 09:26 AM rgw Bug #2369 (Resolved): rgw: bucket attr update does not propagate correctly to all rgw instances
- There were two different issues, both are fixed now:
1. Directly using un-normalized rgw_obj with the cache (Fixed... - 03:23 AM rbd Feature #1484: libvirt: map rbd via kernel driver
- In the current design of libvirt I don't see how you could achieve this.
With my storage pool work I found out tha... - 03:21 AM Feature #1422: libvirt: rbd storage pool
- I just submitted my patch to the libvirt mailinglist. This is a revised version of the patch and should hopefully mak...
05/13/2012
- 03:03 AM Revision f2380f4d (ceph): cephfs: pass -1 for old preferred_osd field
- Signed-off-by: Sage Weil <sage@inktank.com>
05/12/2012
- 09:18 PM Revision bb74b8b3 (ceph): osdmap: filter out nonexistent osds from map
- It is possible that the crush map contains device ids that do not exist as
osds. Filter them out of the CRUSH result... - 04:38 PM Bug #2390: dencoder: depends on expat
- Hmm, ideally ceph-dencoder would still understand the types but not need expat to compile. Which is shouldn't, in pr...
- 04:15 PM Feature #2305 (Resolved): Moving rbd images between pools
- commit:ee26c5d73a48b64292d16a87ebe69908c142048e
- 03:53 PM Bug #2396 (Fix Under Review): don't fail on mismatched size CRUSH and OSD maps
- wip-osdmap
- 03:53 PM Bug #2397: mon: prevent addition of CRUSH items past the max_osd
- wip-osdmap
- 03:53 PM Bug #2397 (Fix Under Review): mon: prevent addition of CRUSH items past the max_osd
05/11/2012
- 11:46 PM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- Hi Josh,
should've been done, if I missedt s/t, stuff is in /usr/src...
Hope it helps,
Oliver.
- 09:03 PM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- Here's a log in which the dispatcher never runs again, despite the osd_map_ack and other messages being received.
... - 09:14 AM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- The dispatch thread hasn't run since this point in the log:...
- 09:12 AM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- So the connection with no pipe problem only happens when the monitor is restarted. The issue still happens otherwise ...
- 11:12 PM Bug #2397 (Resolved): mon: prevent addition of CRUSH items past the max_osd
- To make osd addition a little friendlier, we should add checks to make sure that people run "ceph osd create" before ...
- 11:10 PM Bug #2396 (Resolved): don't fail on mismatched size CRUSH and OSD maps
- See osd-crush-resize. Come up with a better fix to apply to master (or next, if that's going).
- 11:22 AM Linux kernel client Bug #2287: rbd: crashes with 10Gbit network and fio
- We used the attached patch to resolve the immediate problem.
But we still see other crashes over the time. I foun... - 11:05 AM Linux kernel client Bug #2395 (Resolved): kernel crash after unmap a rdb device while the cluster is down
- 1) get a ceph cluster running
2) create a RBD and map it to an client
3) shutdown the cluster
4) as soon as the cl...
05/10/2012
- 05:44 PM Feature #2394 (Resolved): Provide tool to answer: "when is it safe to kill this osd"
- After "ceph osd out 123", when is it safe to kill the ceph-osd daemon?
Assume a busy cluster where there's other f... - 05:10 PM Revision 80bbe5e4 (ceph): keyring: make child of KeyStore
- This lets us pass a keyring to the auth methods as a source for keys for
doing the authentication handshaking. Norma... - 05:09 PM Revision 7ce157d6 (ceph): utime_t: no double ctor
- error: os/FileJournal.h:48:51: call of overloaded ‘utime_t(int)’ is ambiguous
Signed-off-by: Sage Weil <sage@newdrea... - 04:13 PM Bug #2393: objecter: dropping messages (old connection being used)
- Ah, too bad.
- 03:55 PM Bug #2393: objecter: dropping messages (old connection being used)
- The caller always holds the lock for the objecter. RadosClient::ms_handle_reset grabs the lock and calls objecter->ms...
- 01:49 PM Bug #2393: objecter: dropping messages (old connection being used)
- Objecter::ms_handle_reset() was not acquiring a lock. ms_handle_reset() racing with any operation that grabs the sess...
- 10:57 AM Bug #2393: objecter: dropping messages (old connection being used)
- Ah, need to look at the logs again. There's nothing in this excerpt to say that both requests were supposed to go to ...
- 03:39 PM Bug #2316: rbd: restart of OSD leeds to stale qem-VM's with "ceph version 0.45-207-g3053e47"
- With more debugging, this looks like a problem with the messenger or the monitor client's use of it. After a fault, t...
Also available in: Atom