Activity
From 07/16/2012 to 08/14/2012
08/14/2012
- 11:54 PM Revision 36d6c4f4 (ceph): Merge branch 'next'
- 11:53 PM Revision 119880ca (ceph): Merge branch 'testing' into next
- 11:53 PM Revision 24a26c62 (ceph): mkcephfs: fix mon_data check
- * check the right path <facepalm>
* behave if the directory doesn't exist at all yet
Fixes: #2922
Signed-off-by: Sag... - 11:17 PM Revision bcd00c65 (ceph): Merge branch 'wip-2918'
- Reviewed-by: Dan Mick <dan.mick@inktank.com>
- 11:06 PM Revision 3d3d91df (ceph): osdmap: apply mon_max_osd when generating osdmap from conf
- This prevents users from having an [osd.1234567] section and blowing up
their memory usage.
Signed-off-by: Sage Weil... - 11:06 PM Revision 1a5e12ec (ceph): osdmap: fix pg_num calculation when generating osdmap from conf
- Base num_pg calculation on the number of osds, not the max osd id.
Fixes: #2918
Signed-off-by: Sage Weil <sage@inkta... - 10:08 PM Revision 6dbbcf03 (ceph): queue: fix logging of child return code
- 10:07 PM Revision 615f85d0 (ceph): mon: throttle daemon messages independently from client messages
- This lets us set a higher limit on messages from daemons (osd, mds) than
from clients, and throttle them independentl... - 09:59 PM Revision cca85af5 (ceph): global: only print banner if type is daemon
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:45 PM Revision 2f3e03f6 (ceph): perfcounters: add u64 average type
- Signed-off-by: Sage Weil <sage@inktank.com>
- 09:45 PM Revision 9fc79584 (ceph): filejournal: instrument journal write counts, sizes
- This rounds out the low-hanging fruit that are coming to mind. Moving on.
Fixes: #2619
Signed-off-by: Sage Weil <sa... - 09:06 PM Revision 1eea43db (ceph): global: fix version banner in log on startup
- Log banner from global_init(). This puts it at the very top of the log,
right when the daemon starts.
Fixes: #2940
... - 08:13 PM Revision efe913b8 (ceph): Merge branch 'wip-crush-tunables'
- Reviewed-by: Greg Farnum <greg@inktank.com>
- 08:11 PM Revision 32671276 (ceph): doc: document use of CRUSH tunables
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:42 PM Revision 73218a7a (ceph): msgr: make set_policy() and set_default_policy() safe to re-use
- Introduce a policy_lock in SimpleMessenger to make this safe even after
the messenger has been started up. The user ... - 05:42 PM Revision da35b4c2 (ceph): msgr: make set_policy_throttler safe, act on default
- Add locking in set_policy_throttler.
Also, make it act on the default policy when the specified type does not
have a... - 05:42 PM Revision 43f7d374 (ceph): osd: put throttler on default policy
- Before we were associating the throttler with a CLIENT specific policy
that didn't cover MDS. That's silly. Instead... - 05:42 PM Revision 3dd5ebdc (ceph): osd: require CRUSH_TUNABLES feature of peers, clients when used
- If the current crush map has non-default tunables, set our messengers to
require the tunables feature bit. If it doe... - 05:42 PM Revision b254ba78 (ceph): mon: require CRUSH_TUNABLES when latest osdmap has tunables set
- This is an imperfect check in that tunables may be set in past osdmaps,
and the OSDs may need that in order to peer c... - 05:06 PM Revision 5ab4939c (ceph): doc: v0.48.1argonaut release notes, changelog
- Signed-off-by: Sage Weil <sage@inktank.com>
- 04:22 PM Revision dc2a232b (ceph): mon: simplify logmonitor check_subs; less noise
- * simple helper to translate name to id
* verify sub type is valid in caller
* assert sub type is valid in method
...
08/13/2012
- 11:18 PM Revision 99ac6b0b (ceph): Disable asynchronous DNS lookups.
- Especially on older hosts, we keep triggering errors::
ServerNotFoundError: Unable to find the server at
teuthol... - 09:58 PM Revision a7ad701b (ceph): v0.48.1argonaut
- 08:34 PM Revision 4e7d49e0 (ceph): Merge branch 'wip-msgr'
- 08:28 PM Revision 75172c70 (ceph): msg/Accepter: fix nonce initialization
- This needs to be provided to the Accepter at bind time, not by start().
Otherwise the nonce is effectively always 0, ... - 06:46 PM Revision 294c25bb (ceph): ceph-osd: log journal-creation failure with derr
- Fixes: #2938
Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 06:03 PM Revision 1740bd0c (ceph): crush: add helper has_nondefault_tunables()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 06:03 PM Revision 8af2cf3c (ceph): msgr: expose get_policy() through generic Messenger API
- Return a copy instead of a const reference, too...
Signed-off-by: Sage Weil <sage@inktank.com> - 04:58 PM Revision 95ecab27 (ceph): Merge branch 'wip-msgr'
- 04:58 PM Revision c8af0faf (ceph): cephtool: make command wait for osdmap explicit
- If we are waiting for an osdmap for a command, track that explicitly so
that we know when to retry send_command().
S... - 04:49 PM Revision a4428bd0 (ceph): v0.50
- 04:48 PM Revision c75c0d81 (ceph): Merge tag 'v0.50'
- v0.50
- 03:46 PM Revision ac02b34e (ceph): msg/Pipe: discard_queue() -> discard_out_queue()
- We only discard outgoing messages; incoming messages are handled by the
IncomingQueue.. but this method doesn't touch... - 03:46 PM Revision 8c1632ba (ceph): cephtool: send keepalive to tell target
- If we 'ceph tell <foo> ...' to a non-monitor, we need to send keepalives to
ensure we detect a tcp drop. (Not so for... - 03:46 PM Revision b97f6e35 (ceph): msgr: some SimpleMessenger docs
- Document basic modules and the lock ordering.
Signed-off-by: Sage Weil <sage@inktank.com> - 03:46 PM Revision 608c776b (ceph): msgr: remove useless SimpleMessenger::msgr
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:46 PM Revision c84b7289 (ceph): msg/Accepter: make members private
- Nobody uses these.
Signed-off-by: Sage Weil <sage@inktank.com> - 03:46 PM Revision a0017fcc (ceph): msgr: protect set_myaddr()
- This is used by Messenger implementation (and their constituent
components).
Signed-off-by: Sage Weil <sage@inktank.... - 03:46 PM Revision 8453a819 (ceph): msg/Accepter: pass nonce on start
- This lets us drop the otherwise awkward SimpleMessenger::get_nonce()
accessor.
Signed-off-by: Sage Weil <sage@inktan... - 03:46 PM Revision 1b8f2e05 (ceph): msg/SimpleMessenger: push need_addr check into learned_addr()
- This puts all of the do/do not lock logic in one place, and documents
it.
Signed-off-by: Sage Weil <sage@inktank.com> - 03:46 PM Revision 5d5045d3 (ceph): msg/Accepter: use learned_addr() from Accepter::bind()
- Normally we never go from need_addr == false to need_addr == true.
It always starts out as true, so this else is usel... - 03:46 PM Revision 76954c13 (ceph): msg/Pipe: document tcp_*()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:46 PM Revision d58df35f (ceph): msg/Pipe: simplify Pipe::tcp_read() return value
- 0 for success; no reason to return length (always == len).
Signed-off-by: Sage Weil <sage@inktank.com> - 03:45 PM Revision b30ad9a1 (ceph): cephtool: fix deadlock on fault when waiting for osdmap
- send_command() was blocking for the osdmap, and also called from the
connect callback. Instead, re-call it from the ... - 03:45 PM Revision ee206a52 (ceph): cephtool: set messenger policy
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:45 PM Revision 90e0ef90 (ceph): cephtool: retry 'ceph tell <who> ...' command if connection fails
- It was easy to reproduce a hang with 'ceph osd tell osd.0 foo' and
messenger failure injection.
Signed-off-by: Sage ...
08/12/2012
- 11:42 PM Revision ce0bc11a (ceph): rbd: show snap protection status
- If you do rbd info image@snap, include whether that snapshot is protected.
Signed-off-by: Josh Durgin <josh.durgin@i...
08/10/2012
- 05:39 PM Revision d4849f2f (ceph): rgw: fix usage trim call encoding
- Fixes: #2841.
Usage trim operation was encoding the wrong op structure (usage read).
Since the structures somewhat ov... - 05:39 PM Revision 515952d0 (ceph): cls_rgw: fix rgw_cls_usage_log_trim_op encode/decode
- It was not encoding user, adding that and reset version
compatibility.
This changes affects command interface, makes ... - 05:39 PM Revision 2e77130d (ceph): rgw: expand date format support
- Relaxing the date format parsing function to allow UTC
instead of GMT.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 05:39 PM Revision 14fa77d9 (ceph): rgw: complete multipart upload can handle chunked encoding
- Fixes: #2878
We now allow complete multipart upload to use chunked encoding
when sending request data. With chunked e... - 05:39 PM Revision a06f7783 (ceph): rgw_xml: xml_handle_data() appends data string
- Fixes: #2879.
xml_handle_data() appends data to the object instead of just
replacing it. Parsed data can arrive in pi... - 05:39 PM Revision a8b224b9 (ceph): rgw: ETag is unquoted in multipart upload complete
- Fixes #2877.
Removing quotes from ETag before comparing it to what we
have when completing a multipart upload.
Backp... - 03:59 PM Revision 98286b49 (ceph): MonMap: return error on failure in build_initial
- If mon_host fails to parse, return an error instead of success.
This avoids failing later on an assert monmap.size() ... - 03:59 PM Revision 5c116bf8 (ceph): addr_parsing: report correct error message
- getaddrinfo uses its return code to report failures.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 03:58 PM Revision 22259c6e (ceph): MonMap: return error on failure in build_initial
- If mon_host fails to parse, return an error instead of success.
This avoids failing later on an assert monmap.size() ... - 03:58 PM Revision 49b2c7b5 (ceph): addr_parsing: report correct error message
- getaddrinfo uses its return code to report failures.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
08/09/2012
- 11:41 PM Revision 1fd7acd8 (ceph): rbd: fix test compilation on 32-bit
- 32-bit can't implicitly convert size_t or unsigned to uint64_t
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 10:12 PM Revision be6b5e1b (ceph): test_librbd_fsx: fix warning
- fstat was included implicitly before.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 09:53 PM Revision 4d71ca63 (ceph): rbd: fix cli tests
- add flatten, snap protect, and snap unprotect
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 08:54 PM Revision 677934b1 (ceph): rgw: modify some error messages
- Make them more correct, clearer.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 08:39 PM Revision 86be6fe7 (ceph): Merge remote-tracking branch 'origin/wip-rgw-admin'
- 08:33 PM Revision b28db08e (ceph): rgw: use multiple notification objects
- Issue #2504. This makes us listen and notify on more than
a single object, which reduces the contention of cache
noti... - 08:27 PM Revision b15d6392 (ceph): radosgw-admin.rst: Misc doccumentation update.
- Signed-off-by: caleb miles <caleb.miles@inktank.com>
- 08:27 PM Revision 45f7f060 (ceph): rgw_admin.cc: Allow removal of a user's buckets during user removal.
- Allow the buckets, and any child objects, of a user to be deleted when the
user is deleted through radosgw-admin. In ... - 06:14 PM Revision c77f0fb7 (ceph): rbd: add snap [un]protect commands
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 06:14 PM Revision cd5d7241 (ceph): librbd: fix memory leak on error in clone
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 06:14 PM Revision 53d77502 (ceph): librados: add method to get the fsid of a cluster
- This will be used by OpenStack to check whether two components
have access to the same backend cluster.
Signed-off-b... - 06:14 PM Revision 8d5f1e93 (ceph): librbd: add methods for protecting/unprotecting snapshots
- Unprotect will be more useful once integrated with the rbd_children object.
Right now we just check that a snapshot i... - 04:55 PM Revision cc8eac24 (ceph): rgw_admin.cc: Allow for deletion of objects through radosgw-admin.
- Allow objects to be deleted through radosgw-admin with an optional flag
to delete the tail of that object during the ... - 04:42 PM Revision 273a43ed (ceph): Flush data to temp file before reading it in another process.
- 04:42 PM Revision 8aaf21d5 (ceph): Oops tempfile now gives us file objects not fds.
- 04:21 PM Revision f9359f0c (ceph): doc: New example usage.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
08/08/2012
- 11:09 PM Revision 6bc1067f (ceph): rgw: fix usage trim call encoding
- Fixes: #2841.
Usage trim operation was encoding the wrong op structure (usage read).
Since the structures somewhat ov... - 11:05 PM Revision 04a0eacd (ceph): cls_rgw: fix rgw_cls_usage_log_trim_op encode/decode
- It was not encoding user, adding that and reset version
compatibility.
This changes affects command interface, makes ... - 10:44 PM Revision 3499c0e5 (ceph): cls_rbd: add methods for getting/setting protection status
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 10:37 PM Revision 074c3c0f (ceph): rgw: expand date format support
- Relaxing the date format parsing function to allow UTC
instead of GMT.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 10:34 PM Revision d39ea1d4 (ceph): rgw: complete multipart upload can handle chunked encoding
- Fixes: #2878
We now allow complete multipart upload to use chunked encoding
when sending request data. With chunked e... - 10:34 PM Revision 03b787e0 (ceph): rgw_xml: xml_handle_data() appends data string
- Fixes: #2879.
xml_handle_data() appends data to the object instead of just
replacing it. Parsed data can arrive in pi... - 10:34 PM Revision 3809e344 (ceph): rgw: ETag is unquoted in multipart upload complete
- Fixes #2877.
Removing quotes from ETag before comparing it to what we
have when completing a multipart upload.
Backp... - 10:06 PM Revision 52f03dcc (ceph): doc: Added debug and logging reference.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 09:48 PM Revision 99e99758 (ceph): In teuthology-worker, shuffle the child stdout/stderr into our log.
- Otherwise, child can suffer a failure that does not get logged by
it's own exception handling machinery, and we have ... - 09:45 PM Revision 05007f7e (ceph): Minimize scope of try-except.
- os.write and list.append won't raise CalledProcessError, and now
we don't need to try to contain them for temp file c... - 09:44 PM Revision 4b9e1762 (ceph): Use tempfile.NamedTemporaryFile instead of mkstemp.
- Simpler code, no manual cleanup needed. We see a littering of
zero-length temp files from teuthology-worker, and this... - 09:07 PM Revision d78dfe5d (ceph): mkcephfs: use default osd_data, _journal values
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 09:07 PM Revision 3c90ff4e (ceph): mkcephfs: use new default keyring locations
- The ceph-conf command only parses the conf; it does not apply default
config values. This breaks mkcephfs if values ... - 09:01 PM Revision 7084f295 (ceph): mkcephfs: use default osd_data, _journal values
- Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Greg Farnum <greg@inktank.com> - 09:01 PM Revision 96b1a496 (ceph): mkcephfs: use new default keyring locations
- The ceph-conf command only parses the conf; it does not apply default
config values. This breaks mkcephfs if values ... - 08:05 PM Revision 18391059 (ceph): Merge remote-tracking branch 'gh/next'
- 03:15 PM Revision e4e4882d (ceph): mon: differentiate cluster and client auth requirements
- - require 'cluster required' of osds, mdss, and other mons
- require 'service required' of clients
Signed-off-by: Sa... - 03:15 PM Revision e5368891 (ceph): monclient: differentiate clients vs daemons
- If we are a daemon, require 'auth cluster required'. Otherwise, use the
client requirement.
As always, fall back to... - 03:15 PM Revision 5e13fd67 (ceph): keyring: check client requirement list too when loading keyring
- Check client requirements when loading a keyring from the cct. This is
only used to succeed when no keyring is found... - 03:15 PM Revision ac9d429d (ceph): keyring: simplify from_ceph_context() arguments
- Allocate the KeyRing in the caller, and just pass a pointer.
Signed-off-by: Sage Weil <sage@inktank.com> - 03:15 PM Revision 35e79f23 (ceph): keyring: simplify from_ceph_context()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:15 PM Revision 69130d7c (ceph): monclient: move keyring requirement where it belongs
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:15 PM Revision ece78c86 (ceph): doc: some basic cephx documentation
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:15 PM Revision c03f744e (ceph): keyring: make from_ceph_context() a member
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:09 PM Revision d5704bd7 (ceph): buffer: make release() private
- This should only be called by ~ptr or when we are replacing the current
target with something new. It is not suitabl... - 02:33 AM Revision bf4c5e7e (ceph): mds, osd: require auth_service_required, if defined
- Signed-off-by: Sage Weil <sage@inktank.com>
08/04/2012
- 01:49 AM Revision 4fd393f3 (ceph): librbd: fix bounds of memcpy in info()
- Caught by valgrind.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 01:49 AM Revision d48d1ffb (ceph): test_librbd_fsx: add cloning
- 5% of the time a clone op is chosen, create a snapshot, save the
current contents to a file, and continue operating o... - 01:48 AM Revision 90dc5650 (ceph): librbd: implement read/write from parent images
- Put the completion handling logic into new subclases of
librbd::AioRequest, so the caching/non-caching paths can shar... - 01:48 AM Revision beccac93 (ceph): librbd: don't open parent again during get_parent_info
- The parent was already opened when the original ImageCtx
was initialized. We just need to lookup the pool and image
n... - 01:48 AM Revision 12c9a7d5 (ceph): test_rbd.py: refactor cloning tests
- Move into a separate class that requires layering to be enabled,
so the common step of creating and deleting a clone ... - 01:48 AM Revision 965271db (ceph): test_librbd: add a simple io to clone test
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 01:48 AM Revision 53947ed8 (ceph): librbd: check new snapc for validity, not the old one
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 01:48 AM Revision 9bf3d486 (ceph): librbd: move flush on new snap outside of snap_lock
- snap_lock needs to be taken during writeback.
This is still protected by md_lock. The altered snapc doesn't
affect in... - 01:48 AM Revision 12e20709 (ceph): librbd: close parent after shutting down cache
- Requests that haven't been flushed yet might need the parent still.
Signed-off-by: Josh Durgin <josh.durgin@inktank.... - 01:48 AM Revision 745783c1 (ceph): librbd: expose feature bits through librbd.h
- Move them to a separate header so we don't expose
rbd_types.h to the librbd user.
Signed-off-by: Josh Durgin <josh.d... - 12:59 AM Revision 43291951 (ceph): rados.py: add binding for rados_trunc
- Fixes: #2909
Signed-off-by: Evan Felix <evan.felix@pnnl.gov>
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
08/02/2012
- 11:03 PM Revision 3671a37e (ceph): doc: Explicitly state mount.ceph two first arguments are fixed.
- I've hurt myself too many times with this.
Signed-off-by: Tommi Virtanen <tv@inktank.com> - 11:03 PM Revision 17ad4e18 (ceph): doc: mount.ceph is also needed for cephx.
- Signed-off-by: Tommi Virtanen <tv@inktank.com>
- 10:56 PM Revision c961a20e (ceph): doc: Fix toctree structure for man obsync(1).
- Signed-off-by: Tommi Virtanen <tv@inktank.com>
- 08:31 PM Revision 5db3a9e7 (ceph): rgw_admin.cc: Disallow addition of S3 keys with subuser creation
- Fixes: #1855
It is no longer possible to create a subuser and new S3 key associated
with that user through the radosg... - 08:02 PM Revision 4e40a784 (ceph): ceph-authtool: Fix usage, it's --print-key not --print.
- Signed-off-by: Tommi Virtanen <tv@inktank.com>
- 06:14 PM Revision 8665bdc1 (ceph): task: scrub OSDs periodically
- Signed-off-by: Mike Ryan <mike.ryan@inktank.com>
- 06:14 PM Revision 3b85b231 (ceph): task: verify scrub detects files whose contents changed
- Signed-off-by: Mike Ryan <mike.ryan@inktank.com>
- 06:00 PM Revision 25de5e59 (ceph): Revert "osd: peering: detect when log source osd goes down"
- This reverts commit 0d3d75ec0d17e2f6bc5d42058f51eb2ff19ffb71.
Accidental double push. - 03:35 PM Revision 93cbab73 (ceph): doc: Simplify submodules explanation.
- ``git clone --recursive`` does ``init`` & ``update`` for us. Also
avoids incorrect language; there never were submodu... - 03:35 PM Revision 48de9b5b (ceph): doc: Say what to do if submodules are out of date.
- Signed-off-by: Tommi Virtanen <tv@inktank.com>
- 03:35 PM Revision 203dffad (ceph): doc: cd to repository before running git status.
- Signed-off-by: Tommi Virtanen <tv@inktank.com>
- 03:33 PM Revision 6af560de (ceph): doc: Correct Git URL for clone
- Using git@github.com would mean SSHing in, and your average open
source consumer probably won't have a Github account... - 12:17 AM Revision 0d3d75ec (ceph): osd: peering: detect when log source osd goes down
- The Peering state has a generic check based on the prior set osds that
will restart peering if one of them goes down ...
08/01/2012
- 11:34 PM Revision 4bd466d6 (ceph): osd: peering: detect when log source osd goes down
- The Peering state has a generic check based on the prior set osds that
will restart peering if one of them goes down ... - 04:55 PM Revision ca2c3813 (ceph): osd: peering: detect when log source osd goes down
- The Peering state has a generic check based on the prior set osds that
will restart peering if one of them goes down ... - 03:37 AM Revision c2ff66eb (ceph): crank up pjd debugging
07/31/2012
- 10:49 PM Revision bb6e0d0e (ceph): wireshark: update patch
- Update to latest source tree (svn 43768).
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 09:26 PM Revision deec81b4 (ceph): ReplicatedPG: clear waiting_for_ack when we send the commit
- Otherwise, we might send the ack anyway later, after a subsequent
commit is sent resulting in an out-of-order op.
Th... - 08:51 PM Revision e0e72d78 (ceph): Merge remote-tracking branch 'upstream/wip-leveldb-iterators'
- 06:55 PM Revision 87defa88 (ceph): rbd: fix off-by-one error in key name
- Fixes: #2846
Signed-off-by: Sylvain Munaut <tnt@246tNt.com> - 06:54 PM Revision 37d5b462 (ceph): secret: return error on empty secret
- Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
- 06:48 PM Revision f22b95da (ceph): rbd: fix off-by-one error in key name
- Fixes: #2846
Signed-off-by: Sylvain Munaut <tnt@246tNt.com> - 06:48 PM Revision e775ce52 (ceph): secret: return error on empty secret
- Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
- 04:55 PM Revision cd0b5278 (ceph): separate regression suite into topical categories rados, rbd, fs
07/30/2012
- 08:43 PM Revision cda5e8e0 (ceph): PG,ReplicatedPG: clarify scrub state clearing
- scrub_clear_state takes care of clearing the SCRUB and REPAIR
flags. Thus, PG::scrub() needn't clear them again sinc... - 08:39 PM Revision 6d464a21 (ceph): PG::mark_clean(): queue_snap_trim if snap_trimq is not empty
- Currently, we won't queue for snap trim until the next map
update.
Noticed while reviewing another patch, this would... - 08:38 PM Revision 1041b92c (ceph): ReplicatedPG::snap_trimmer: requeue if scrub_block_writes
- Otherwise, we do not continue snap_trimming once scrub is
complete.
Noticed while revewing another patch. This woul... - 06:19 PM Revision 9071764a (ceph): qa: remove rbd image when done
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 06:19 PM Revision 4111501a (ceph): qa: add script to run rbd/librbd tests
- Run on old and new style images, with different features. This is
intended to ease development, as opposed to being p... - 06:19 PM Revision 902bbd79 (ceph): cls_rbd: don't duplicate encoding
- Just skip the parts before encoding instead of returning early.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 06:19 PM Revision 72dc665b (ceph): librbd: make ImageCtx methods take snap_id parameters
- This makes it easier to use without racing with snap_set.
Requests in the cache, for example, store their snap_id
and... - 06:19 PM Revision bd119908 (ceph): librbd: add method to get the size of an image
- To go with the other individual properties instead of stat,
add a method that only gets the size.
Signed-off-by: Jos... - 06:19 PM Revision ea75d1ff (ceph): librbd: check for new snapshots in snap_set
- This allows callers to create a snapshot and then read from
it immediately. No one uses this other than the tests,
fo... - 06:19 PM Revision ff2a96c7 (ceph): rbd: fix copy-pasted snap remove error message
- Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
- 06:19 PM Revision c6bc3e10 (ceph): librbd: refactor synchronous I/O
- Write in terms of the asynchronous functions, so all the logic
is not duplicated. Now there's only a single point whe... - 06:19 PM Revision d3b5ee75 (ceph): librbd: split into multiple files
- This was long overdue. There's still a bunch of stuff in internal.cc
that should be moved elsewhere, but this is a go... - 06:17 PM Revision bbc2a4e1 (ceph): librbd: round #blocks in image_info() for edge cases
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 06:17 PM Revision 237f2409 (ceph): rbd: flatten command (copyup an entire image)
- Fixes: #2561
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 06:17 PM Revision 26ed3841 (ceph): librbd, rbd.py, test_rbd.py: flatten methods
- librbd:
- copyup_block for individual blocks (for read/write path)
- flatten for iterating across whole image (for rb... - 05:49 PM Revision 48038002 (ceph): Merge branch 'wip-osd'
- Reviewed-by: Samuel Just <sam.just@inktank.com>
- 05:49 PM Revision 9e5d4e61 (ceph): osd: initialize send_notify on pg load
- When the PG is loaded, we need to set send_notify if we are not the
primary. Otherwise, if the PG does not go throug... - 05:49 PM Revision f9ff8dd3 (ceph): osd: replace STRAY bit with bool
- We were setting a bit in pg->state that is private to the non-primary
PG. The other bits get shared with the mon etc... - 05:47 PM Revision fb1d5495 (ceph): os: KeyValueDB: re-implement (prefix) iter in terms of whole-space iter
- In-a-nutshell-version: Create a whole-space iterator interface, and
implement the already existing, prefix-based iter... - 05:47 PM Revision 9dd8a333 (ceph): os: KeyValueDB: implement snapshot iterators
- Create a set of functions, to be implemented by derivative classes of
KeyValueDB, responsible for returning an iterat... - 05:47 PM Revision 8e404772 (ceph): test: test_keyvaluedb_iterators: Test KeyValueDB implementations iterators
- This set of tests focus on testing the expected behavior of LevelDBStore's
and KeyValueDBMemory's iterators.
We test... - 02:27 AM Revision 4a0a7e2f (ceph): auth: (ordered) list of auth methods, not a set
- This lets us express a preference.
Signed-off-by: Sage Weil <sage@newdream.net> - 02:27 AM Revision ec6ecc19 (ceph): auth: AuthSupported -> AuthMethodList
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:27 AM Revision effdec9e (ceph): auth: introduce cluster, service, and client auth settings
- Signed-off-by: Sage Weil <sage@inktank.com>
07/29/2012
- 07:40 PM Revision e4e239e2 (ceph): kernel: push a local .deb instead of using gitbuilder
- This lets you specify a path to an existing kernel deb
to be pushed and installed on the remote node.
Limitations:
... - 07:15 PM Revision 1c93d5ab (ceph): syslog check: fix false-positive BUG matches in random strings
07/28/2012
- 08:53 PM Revision 9d43c8a4 (ceph): test: workloadgen: Don't linearly iterate over a map to obtain a collec...
- We were iterating over the collections map a certain amount of times, in
order to obtain the collection in that posit... - 05:54 PM Revision 85187edd (ceph): add osd-recovery-incomplete
- 05:53 PM Revision a0847694 (ceph): osd_recovery: also test unfound discovery
- This tests for bug #2866.
- 05:23 PM Revision 731d5209 (ceph): ceph_manager: count 'incomplete' as 'down'
- 05:23 PM Revision a9f2bf62 (ceph): ceph_manager: wait_for_active
- 05:23 PM Revision 8dd09cb2 (ceph): osd_recovery: test incomplete pg recovery
- 4-osd test to reproduce #2860 and confirm the fix.
- 05:05 PM Revision 7b9d37c6 (ceph): osd: set STRAY on pg load when non-primary
- The STRAY bit indicates that we should annouce ourselves to the primary,
but it is only set in start_peering_interval... - 04:04 PM Revision bae83701 (ceph): osd: peering: make Incomplete a Peering substate
- This allows us to still catch changes in the prior set that would affect
our conclusions (that we are incomplete) and... - 04:04 PM Revision d1602ee2 (ceph): osd: peering: move to Incomplete when.. incomplete
- PG::choose_acting() may return false and *not* request an acting set change
if it can't find any suitable peers with ... - 02:39 PM Revision d6126940 (ceph): config: send warnings to a ostream* argument
- We shouldn't always send these to stderr. (Among other things, the
warning: prefix breaks the gitbuilder error detec... - 02:21 PM Revision 1fc19df8 (ceph): Merge remote-tracking branch 'gh/wip-msgr-masterbits'
- Reviewed-by: Greg Farnum <greg@inktank.com>
07/27/2012
- 11:03 PM Revision 96feca45 (ceph): osd: peering: make Incomplete a Peering substate
- This allows us to still catch changes in the prior set that would affect
our conclusions (that we are incomplete) and... - 10:39 PM Revision a71e442f (ceph): osd: peering: move to Incomplete when.. incomplete
- PG::choose_acting() may return false and *not* request an acting set change
if it can't find any suitable peers with ... - 09:28 PM Revision de4474ac (ceph): vstart.sh: apply extra conf after the defaults
- This let's you do e.g., -o 'debug ms = 100' and it will apply after
the default logging levels.
Signed-off-by: Sage ... - 09:00 PM Revision 623026d9 (ceph): Merge remote-tracking branch 'gh/stable' into stable-next
- 08:42 PM Revision 0b8b58ff (ceph): fix adminsocket test
- 05:45 PM Revision bbc49179 (ceph): msg/Pipe: if we send a wait, make sure we follow through
- Mark our outgoing connection attempt if we send a WAIT in accept(). This
ensures we don't go to standby or closed in... - 05:44 PM Revision 6c01d46e (ceph): client: handle fault during session teardown
- We may have a sequence like:
- client does REQUEST_CLOSE
- mds sends reply
- connection faults, client does get r... - 05:44 PM Revision a879425b (ceph): msg/Pipe: make STANDBY behavior optional
- In particular, lossless_peers should use STANDBY, but lossless_clients
should reconnect immediately since they are al... - 05:43 PM Revision f69d025b (ceph): conf: make dup lines override previous value
- If you put
[some section]
foo = 1
...
foo = 2
in a .conf file, make the second key override the first.
Generate... - 05:43 PM Revision 4788567e (ceph): qa: simple test for pool create/delete commands
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:43 PM Revision 6f7837a9 (ceph): mon: make 'osd rm ...' idempotent
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:43 PM Revision 640e5fde (ceph): qa: simple tests for 'ceph osd create|rm' commands
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:43 PM Revision 3715d205 (ceph): mon: check for invalid pool snap creates in preprocess_op, too
- This avoids waiting for a paxos commit just to return an error.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:43 PM Revision 507f99e9 (ceph): librados: make snap create/destroy handle client-side errors
- Signed-off-by: Sage Weil <sage@inktank.com>
- 05:43 PM Revision 53aa9593 (ceph): objecter: return ENOENT/EEXIST on pool snap delete/create
- Do these checks on the client to mask monitor idempotency from the user.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:43 PM Revision 5f3ef77d (ceph): mon: make pool snap creation ops idempotent
- Return 0 if the snap already exists, or is already deleted.
Also, avoid updating the pg_pool if we are just waiting ... - 05:43 PM Revision 4dfc14c4 (ceph): mon: remove superfluous "can't delete except on master" comments
- That's what 'return false' means for preprocess_*().
Signed-off-by: Sage Weil <sage@inktank.com> - 05:43 PM Revision 6ec9555b (ceph): cond: reorder asserts
- Make the more specific checks assert before the less specific ones, so we
are more likely to crash with useful inform... - 05:43 PM Revision c2e1c629 (ceph): mutex: assert we are unlocked by the same thread that locked
- This only works for non-recursive locks. (Which is probably all of them?)
Signed-off-by: Sage Weil <sage@inktank.com> - 05:43 PM Revision 46e819ed (ceph): objecter: fix mon command resends
- The monitor session is lossy. Send these when the op is initiated, or
when we reconnect. The timeout/cutoff was pre... - 05:43 PM Revision 358d6b61 (ceph): librados: make pool create/destroy handle client-side errors
- Add tests!
Signed-off-by: Sage Weil <sage@inktank.com> - 05:43 PM Revision 597f14ab (ceph): objecter: make pool create/delete return EEXIST/ENOENT
- Do these checks on the client side to mask monitor idempotency from
the user.
Signed-off-by: Sage Weil <sage@inktank... - 05:43 PM Revision 5503376f (ceph): mon: make pool removal idempotent
- Return success if pool does not exist. Part of #2638.
Signed-off-by: Sage Weil <sage@inktank.com> - 05:43 PM Revision a01e22d2 (ceph): mon: make pool creation idempotent
- Return success if the pool already exists. Part of #2638.
Also, fix this so we wait until a creating pool is create... - 04:55 AM Revision 9553c6ed (ceph): osd: fixing sharing of past_intervals on backfill restart
- We need to share past_intervals whenever we instantiate the PG on a peer.
In the PG activation case, this is based on... - 04:55 AM Revision 29aa1cf4 (ceph): filestore: check for EIO in read path
- Check for EIO in read methods and helpers. Try to do checks in low-level
methods (e.g., lfn_*()) to avoid duplicatio... - 04:29 AM Revision 0891948e (ceph): filestore: add 'filestore fail eio' option, default true
- By default we will assert/fail/crash on EIO from the underlying fs. We
already do this in the write path, but not th...
07/26/2012
- 11:35 PM Revision 9db78090 (ceph): osd: fixing sharing of past_intervals on backfill restart
- We need to share past_intervals whenever we instantiate the PG on a peer.
In the PG activation case, this is based on... - 11:29 PM Revision aef10e72 (ceph): librbd: fix id initialization in new format
- 48bd839b1e25b063c675416a8f6233463f1af115 should have included this.
I misread it due to the use of bid instead of id ... - 10:29 PM Revision 17bb78a2 (ceph): librbd: fix id initialization in new format
- 48bd839b1e25b063c675416a8f6233463f1af115 should have included this.
I misread it due to the use of bid instead of id ... - 10:26 PM Revision 5601ae27 (ceph): mon: set a configurable max osd cap
- Don't allow setting a higher osd num through the
ceph control util.
Fixes: #2752
Signed-off-by: Yehuda Sadeh <yehuda... - 10:04 PM Revision 87b6e804 (ceph): Merge remote-tracking branch 'gh/wip-rbd-bid' into stable-next
- 10:04 PM Revision 81d72e5d (ceph): mon: make 'ceph osd rm ...' wipe out all state bits, not just EXISTS
- This ensures that when a new osd reclaims that id it behaves as if it were
really new.
Backport: argonaut
Signed-off... - 10:03 PM Revision ad9c37f2 (ceph): test_stress_watch: just one librados instance
- This was creating a new cluster connection/session per iteration, and
along with it a few service threads and sockets... - 10:03 PM Revision c60afe18 (ceph): Merge commit '35b13266923f8095650f45562d66372e618c8824' into stable-next
- First batch of msgr fixes.
- 10:03 PM Revision d2d40dc3 (ceph): client: fix locking for SafeCond users
- Need to wait on flock, not client_lock.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:03 PM Revision 6ed01df4 (ceph): workqueue: kick -> wake or _wake, depending on locking
- Break kick() into wake() and _wake() methods, depending on whether the
lock is already held. (The rename ensures tha... - 10:03 PM Revision 7fbbe465 (ceph): librados: take lock when signaling notify cond
- When we are signaling the cond to indicate that a notify is complete,
take the appropriate lock. This removes the po... - 10:03 PM Revision 4d7d3e27 (ceph): osd: guard class call decoding
- Backport: argonaut
Signed-off-by: Sage Weil <sage@inktank.com> - 10:03 PM Revision 682609a9 (ceph): objecter: always resend linger registrations
- If a linger op (watch) is sent to the OSD and updates the object, and then
the client loses the reply, it will resend... - 10:03 PM Revision 88330505 (ceph): ReplicatedPG: fix replay op ordering
- After a client reconnect, the client replays outstanding ops. The
OSD then immediately responds with success if the ... - 10:02 PM Revision c963a21a (ceph): filestore: check for EIO in read path
- Check for EIO in read methods and helpers. Try to do checks in low-level
methods (e.g., lfn_*()) to avoid duplicatio... - 10:01 PM Revision 6bd89aeb (ceph): filestore: add 'filestore fail eio' option, default true
- By default we will assert/fail/crash on EIO from the underlying fs. We
already do this in the write path, but not th... - 08:48 PM Revision 0d6ce424 (ceph): Fixed the code to pass 'yes' during mkfs
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 01:13 AM Revision e9b5a289 (ceph): config: fix 'config set' admin socket command
- Fixes: #2832
Backport: argonaut
Signed-off-by: Sage Weil <sage@inktank.com>
07/25/2012
- 11:38 PM Revision 2b75ddef (ceph): Added '-y' option for mkfs.ext4
- Signed-off-by: tamil <tamil.muthamizhan@inktank.com>
- 11:35 PM Revision 1a6cd965 (ceph): osd: break potentially large transaction into pieces
- We do a similar trick elsewhere. Control this via a tunable. Eventually
we'll control the others (in a non-stable b... - 10:57 PM Revision bcb9ab8b (ceph): doc: updates to fix problem with ceph-cookbooks appearing in chef-server.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 09:53 PM Revision 15e16229 (ceph): osd: only commit past intervals at end of parallel build
- We don't check for gaps in the past intervals, so we should only commit
this when we are completely done. Otherwise ... - 08:28 PM Revision 9767146f (ceph): osd: generate past intervals in parallel on boot
- Even though we aggressively share past_intervals with notifies etc, it is
still possible for an osd to get buried beh... - 08:28 PM Revision d45929f4 (ceph): osd: move calculation of past_interval range into helper
- PG::generate_past_intervals() first calculates the range over which it
needs to generate past intervals. Do this in ... - 08:27 PM Revision 18d5fc41 (ceph): osd: fix map epoch boot condition
- We only want to join the cluster if we can catch up to the latest
osdmap with a small number of maps, in this case a ... - 05:58 PM Revision fca65ff5 (ceph): osd: move calculation of past_interval range into helper
- PG::generate_past_intervals() first calculates the range over which it
needs to generate past intervals. Do this in ... - 05:58 PM Revision 16302ace (ceph): osd: generate past intervals in parallel on boot
- Even though we aggressively share past_intervals with notifies etc, it is
still possible for an osd to get buried beh... - 05:58 PM Revision 5979351e (ceph): osd: fix map epoch boot condition
- We only want to join the cluster if we can catch up to the latest
osdmap with a small number of maps, in this case a ... - 04:36 AM Revision 61ff894c (ceph): admin-socket: test generic admin socket commands
- 03:54 AM Revision 11b275a0 (ceph): osd: avoid misc work before we're active
- If we're booting, we shouldn't scrub, or send reports to the montior,
or send heartbeats, or any of that.
Signed-off... - 03:51 AM Revision 278b5f58 (ceph): mon: ignore pgtemp messages from down osds
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:51 AM Revision 08e2ecac (ceph): mon: ignore osd_alive messages from down osds
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:18 AM Revision 8c7186d0 (ceph): mon: ignore pgtemp messages from down osds
- Signed-off-by: Sage Weil <sage@inktank.com>
- 03:16 AM Revision b17f5467 (ceph): mon: ignore osd_alive messages from down osds
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:23 AM Revision 404a7f52 (ceph): admin_socket: json output, always
- If the perfcounters stuff were refactored to use the Formatter, we could
put the JSONFormatter in the admin_socket co... - 12:23 AM Revision 0133392b (ceph): admin_socket: dump config in json; add test
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:22 AM Revision 8c3b4907 (ceph): Merge branch 'next'
07/24/2012
- 10:26 PM Revision 9bc86171 (ceph): admin_socket: make test optional
- If it's not there, we just verify the output is valid json.
- 08:53 PM Revision 0ef8cd3c (ceph): config: fix 'config set' admin socket command
- Fixes: #2832
Backport: argonaut
Signed-off-by: Sage Weil <sage@inktank.com> - 06:49 PM Revision 186a595c (ceph): Merge branch 'next'
- 06:02 PM Revision f565ace6 (ceph): osd: fix pg log zeroing
- Zero the right number of bytes. Fixes a bug where we clobber legit log
data. Fortunately this is only triggered wit... - 05:10 PM Revision 3e886799 (ceph): Merge branch 'wip-2763'
- 05:09 PM Revision 52f51a24 (ceph): wireshar/ceph/packet-ceph.c: fix eol
- Removing extra char from dos eol format.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 05:09 PM Revision d67ad0db (ceph): Wireshark dissector updated, work with the current development tree of ...
- 01:30 AM Revision a16d9c64 (ceph): os: KeyValueDB: allow finer-grained control of transaction operations
- This patch introduces the possibility of using single key/value
modification operations into the transaction interfac... - 01:30 AM Revision a3d57a6e (ceph): os: KeyValueDB: Add virtual raw_key() function to return (prefix,key) pair
- If we were to use solely the key() function, whenever we had a key with,
say, prefix 'Foo' and key 'Bar', the key() f... - 12:39 AM Revision 6c0fa509 (ceph): doc: update information about stable vs development releases
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:16 AM Revision 48bd839b (ceph): librbd: replace assign_bid with client id and random number
- The assign_bid method has issues with replay because it is a write
that also returns data. This means that the replay...
07/23/2012
- 11:51 PM Revision 67832c34 (ceph): osd: fix ACK ordering on resent ops
- The wait_for_ondisk handling fixed COMMIT ordering, but the ACKs need to
go back in the same order too. For example:... - 11:01 PM Revision 96dbc412 (ceph): rados::cls::lock: move api types into namespace
- By popular demand, moved public api into namespace. This
required some changes to ceph_dencoder to get some template
... - 09:09 PM Revision 7dfdf4f8 (ceph): librbd: replace assign_bid with client id and random number
- The assign_bid method has issues with replay because it is a write
that also returns data. This means that the replay... - 07:51 PM Revision dc2d6711 (ceph): librados: add new constructor to form a Rados object from IoCtx
- This creates a separate reference to an existing connection, for
use when a client holding IoCtx needs to consult ano... - 07:43 PM Revision d9bfe954 (ceph): Merge tag 'v0.49'
- v0.49
- 06:28 PM Revision ca6265d0 (ceph): v0.49
- 05:47 PM Revision c8f13119 (ceph): mon: make 'ceph osd rm ...' wipe out all state bits, not just EXISTS
- This ensures that when a new osd reclaims that id it behaves as if it were
really new.
Backport: argonaut
Signed-off... - 04:56 PM Revision f42e1875 (ceph): cls_rbd, cls_rbd_client, test_cls_rbd: copyup method
- Fixes: #2559
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 04:55 PM Revision 9b9efe7e (ceph): cls_rbd_client: fix locking function indentation
- Greg's editor was set to spaces only.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 04:55 PM Revision d5e45418 (ceph): cls_rbd: return negative pool id if parent does not exist
- This lets us use get_parent in a multi-op transaction without getting
-ENOENT for the entire transaction.
Signed-off... - 04:55 PM Revision 56bc3699 (ceph): cls_rbd: make get_parent return valid data when layering is disabled
- This means clients can treat an error in their multi-object
transaction as a failure for all of them. This makes the ... - 04:55 PM Revision 02aba7bc (ceph): librbd: rename variables for more consistency
- Rename: poolname -> pool_name
snapname -> snap_name
snapid -> snap_id
Signed-off-by: Josh Durgin <... - 04:55 PM Revision 4af1f981 (ceph): librbd: move watch/unwatch to ImageCtx methods
- This fixes close_image() on a partially initialized ImageCtx, and
paves the way for opening parent images without wat... - 04:55 PM Revision 233d44a1 (ceph): librbd: allow an image to be opened without watching
- Watching the header of a parent image could produce unreasonable
delays. If hundreds of child images watch the same p... - 04:55 PM Revision df71c2d3 (ceph): librbd: store parent info in snapshot metadata
- If an image is open to a snapshot, return the features or overlap from
that snapshot.
Signed-off-by: Josh Durgin <jo... - 04:55 PM Revision 8014d932 (ceph): librbd: allow opening an image by id
- If we have an ImageCtx with an id set, don't bother setting the name.
Just go directly to the header object in ImageC... - 04:55 PM Revision b82f5958 (ceph): librbd: move all side-effects to snap_{un}set()
- All callers need to do this anyway.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 04:55 PM Revision c18c4b86 (ceph): librbd: open parent image when opening a child
- Extract a helper out of get_parent_info. The parent may become unset
while the child is open, so detect changes in it... - 04:55 PM Revision 127ff618 (ceph): librbd: drop unnecessary std:: and struct prefixes
- Most of this file do not use them.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 04:21 PM Revision 5fcb22f0 (ceph): mkcephfs: add sync between btrfs scan and mount
- This appears to fix problems with mount failing for at least one user.
Reported-by: Paul Pettigrew <Paul.Pettigrew@m... - 05:03 AM Revision d0698f9d (ceph): schedule_suite.sh: put coverage option in ceph section
- 03:50 AM Revision deb7a54a (ceph): marginal kclient+ffsb: enable mds logging to catch badess
- See #1947
- 03:47 AM Revision 32353f13 (ceph): move misc, blogbench back into active kernel suite
- these were removed from regression ages ago, and only recently put back in
marginal. they seem fine. - 03:47 AM Revision 000fec3b (ceph): move all kernel tests to kernel suite; symlink collections from regression
- Make regression a union of other topical suites.
07/22/2012
- 03:59 AM Revision 17e4f751 (ceph): this fails reliably
- 03:18 AM Revision f70b8250 (ceph): ceph: fix mkfs/mount option defaults
- Later code expects a list, not None.
07/21/2012
- 04:15 PM Revision 2d7e2cbf (ceph): crush: fix name map encoding
- We screwed up and encoded using the name 'int' type instead of int32_t.
That means people have systems encoding this ... - 03:50 PM Revision b497bdac (ceph): osd/OpTracker: fix use-after-free
- And formatting.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:36 AM Revision 7cf1f1fb (ceph): msg/Pipe: go to STANDBY on lossless accept fault
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:36 AM Revision 2e67b7a3 (ceph): msgr: do not reopen failed lossy Connections
- There was a race where:
- sending stuff to a lossy Connection
- it fails, and queues itself for reap, queues a RES... - 01:36 AM Revision adce6df2 (ceph): msgr: simplify submit_message()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:36 AM Revision 04fde518 (ceph): msgr: drop CLOSED checks during queueing
- AFAICS these checks are pointless. There should be no harm in queueing
messages on a closed connection; they'll get ... - 01:36 AM Revision 60eb36ef (ceph): msgr: close get_connection() race
- This could null deref if the Pipe is registered but failed.
We need to loop here because the Pipe vs Connection stuf... - 01:36 AM Revision ea7511b8 (ceph): msg/Pipe: fix leak of Connection in ctor
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:36 AM Revision 9348bb38 (ceph): osd: reopen heartbeat connections when they fail
- If we have an active peer whose Connection fails, open a new one. This
is necessary now that a lossy client connecti... - 01:36 AM Revision ef3fd1c3 (ceph): msg/Pipe: go to standby on lossless server connection faults
- Go directly to the STANDBY state, and print a more accurate message.
Otherwise, we do the same check in writer() and ... - 01:36 AM Revision 8966f71a (ceph): msg/IncomingQueue: make the pipe parent informational only
- Use this pointer only for debug output prefix; do not dereference, as we
may live beyond the original parent.
Signed... - 01:36 AM Revision 525830cd (ceph): msgr: do not discard_queue in Pipe reaper
- The IncomingQueue can live beyond the Pipe. In particular, there is no
reason not to deliver messages we've received... - 01:36 AM Revision 7b398a5d (ceph): msgr: indicate whether clients are lossy
- We need to know whether the client is lossy before we connect to the peer
in order to know whether to deliver a RESET... - 01:36 AM Revision 9d94ed1c (ceph): test_stress_watch: verify that the watch operation isn't slow
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:36 AM Revision 472d14f7 (ceph): msg/DispatchQueue: fix locking in dispatch thread
- The locking was awkward with locally delivered messages.. we dropped dq
lock, inq lock, re-took dq lock, etc. We wo... - 01:36 AM Revision cef85105 (ceph): msg/Connection: add failed flag for lossy Connections
- If a lossy Connection fails and we disconnect the Pipe, set a failed flag.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:36 AM Revision 541694f7 (ceph): msg/Pipe: disconnect Pipe from lossy Connection immediately on failure
- When we have a lossy connection failure, immediately disconnect the Pipe
and set the Connection failed flag. There i... - 01:36 AM Revision 9a4e7027 (ceph): msg/Pipe: unregister pipe immediately on fault; fix mark_down
- This fixes a problem where:
- pipe faults, con->pipe is cleared
- ms_handle_reset tries to mark_down, but it doesn... - 01:36 AM Revision 35b7bca3 (ceph): msg: kill tcp.{cc,h}
- Move the remaining comparator into msg_types.h and kill this off.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:36 AM Revision 5a62dfef (ceph): msg/DispachQueue: hold lock in IncomingQueue::discard_queue()
- This prevents races with the dispatch thread, among other things.
Signed-off-by: Sage Weil <sage@inktank.com> - 01:36 AM Revision 999c506d (ceph): msg/DispatchQueue: give IncomingQueue ref to queue
- We want to be able to queue an event (e.g., RESET) and deliver it even
after the Pipe is destroyed.
Signed-off-by: S... - 01:22 AM Revision 5ecc5bce (ceph): msg/DispatchQueue: cleanup debug prefix
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:22 AM Revision 89b07f47 (ceph): msg/Pipe: move tcp_* functions into Pipe class
- This lets us print nice debug prefixes. It also calls BS on the
Pipe vs tcp.cc separation.
Signed-off-by: Sage Weil... - 01:22 AM Revision d034e46d (ceph): msgr: move Accepter into separate .cc
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:22 AM Revision 3e98617c (ceph): msg/Pipe: get_state_name()
- Signed-off-by: Sage Weil <sage@inktank.com>
- 01:22 AM Revision f78a4010 (ceph): msgr: rework accept() connect_seq/race handling
- We change a couple of key things here:
* If there is a matching connect_seq and the existing connection is in OPEN ... - 12:36 AM Revision ec4ce8bf (ceph): regression: do some tests on ext4
- 12:20 AM Revision a6735ab0 (ceph): OpRequest,OSD: track recent slow ops
- This should be helpful while investigating slow performance.
OpRequests now track events with timestamp in addition
...
07/20/2012
- 09:32 PM Revision d624f343 (ceph): Merge branch 'next'
- 08:59 PM Revision 9e207aa8 (ceph): test/store_test.cc: verify collection_list_partial results are sorted
- Synthetic test now also varies snapshots and uses a small variety of
hashes.
Signed-off-by: Samuel Just <sam.just@in... - 08:41 PM Revision 49877cde (ceph): cls_lock: cls_lock_id_t -> cls_lock_locker_id_t
- Renamed type to make more sense.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 08:28 PM Revision 315bbea5 (ceph): cls_lock: document lock properties
- Added some comments about different lock properties.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 08:16 PM Revision 056d42cf (ceph): cls_log: update a comment
- Was missing output param description.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 08:14 PM Revision da770146 (ceph): move cfuse+dbench back to regression for verify, too
- 08:11 PM Revision 2c7d7821 (ceph): rados: lock info keeps expiration, not duration
- We pass duration in the request, but internally we keep
the expiration.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.... - 08:00 PM Revision d16844c8 (ceph): rados tool: add advisory lock control commands
- Can now lock, break lock, list locks and show lock
info.
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com> - 07:59 PM Revision 2f8de894 (ceph): cls_lock: objclass for advisory locking
- Providing an objclass to create and manipulate advisory
locking. Also providing a client api to control it. A lock
ma... - 07:55 PM Revision 9c5c3edf (ceph): objclass: add api calls to get/set xattrs
- added the following functions:
cls_cxx_getxattr
cls_cxx_getxattrs
cls_cxx_setxattr
Signed-off-by: Yehuda Sadeh... - 07:29 PM Revision adc9b91f (ceph): os/HashIndex: use set<pair<string, hobject_t>> rather than multimap
- Multimap does not make any guarantees about ordering of different
values with the same key. list_by_hash, however, a... - 05:14 AM Revision 0b84384f (ceph): mon: shut up about sessionless MPGStats messages
- If the mon gets a reset on the client connection, it clears the session
on the connection. This is perfectly normal ... - 04:27 AM Revision 6580450f (ceph): osd: clean up boot method names
- Prefix subsequent steps with _. Better names.
Signed-off-by: Sage Weil <sage@inktank.com>
Reviewed-by: Yehuda Sadeh... - 04:27 AM Revision 369fbf61 (ceph): osd: defer boot if heartbeatmap indicates we are unhealthy
- If the OSD is bogged down or unresponsive, we should not try to join
the cluster. This was observed on congress (slo... - 03:22 AM Revision d76df212 (ceph): Merge branch 'next'
- Conflicts:
src/include/ceph_features.h - 12:13 AM Revision 7f58b9be (ceph): mon: track pg creations by osd
- Track the pending pg creations by osd, and use a helper to send out that
messages.
Signed-off-by: Sage Weil <sage@in... - 12:13 AM Revision dec93692 (ceph): osd/mon: subscribe (onetime) to pg creations on connect
- Ask the monitor for pending pg creations each time we connect.
Normally, this is a freebie check. If there are pend...
07/19/2012
- 11:48 PM Revision 42de6873 (ceph): rbd: fix usage for snap commands
- Snap commands take '--snap <snapname> <imagename>'.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:45 PM Revision 4c6c927b (ceph): Revert "rbd: fix usage for snap commands"
- This reverts commit 42de6873f9ca33fc20e70176d9a422635a6f0152.
Actually, these are fine! Dan made them all kinds of ... - 10:51 PM Revision e1c98e7d (ceph): tasks: add multibench task for testing pool creation
- Also adds support for specifying a pool for radosbench
to create and then cleanup instead of "data".
Signed-off-by: ... - 06:29 PM Revision 58cd27fd (ceph): doc: add missing dependencies to README
- Signed-off-by: Mike Ryan <mike.ryan@inktank.com>
- 02:50 AM Revision c9967120 (ceph): add CRUSH_TUNABLES feature bit
- Signed-off-by: Sage Weil <sage@inktank.com>
- 02:49 AM Revision 6f381aff (ceph): add CRUSH_TUNABLES feature bit
- Signed-off-by: Sage Weil <sage@inktank.com>
07/18/2012
- 10:37 PM Revision e3349a2a (ceph): OSD::handle_osd_map: don't lock pgs while advancing maps
- We no longer do anything with the pgs here. PG map
advancing is now handled in OSD::advance_pg asyncronously.
Signe... - 09:26 PM Revision ca9f7130 (ceph): OSD: actually send queries during handle_pg_create
- During the osd threading refactor, we lost the do_queries
call in favor of dispatch_context. However, this did not
i... - 09:26 PM Revision 8f5562ff (ceph): OSD: write_if_dirty during get_or_create_pg after handle_create
- In the case that the pg is newly created, we will activate during
that call, so the info and log will be dirty.
Sign... - 09:26 PM Revision c8ee3016 (ceph): osd: add osd_debug_drop_pg_create_{probability,duration} options
- This will let us exercise more of the pg creation code.
Signed-off-by: Sage Weil <sage@inktank.com> - 08:44 PM Revision c49daeca (ceph): clock: print skew with ntp servers to log to help debug time issues
- 07:59 PM Revision 0b579546 (ceph): ObjectCacher: fix cache_bytes_hit accounting
- Misses are not hits!
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 07:58 PM Revision 0d0b4689 (ceph): Merge branch 'next'
- 07:55 PM Revision 5dd68b95 (ceph): objecter: always resend linger registrations
- If a linger op (watch) is sent to the OSD and updates the object, and then
the client loses the reply, it will resend... - 06:04 PM Revision 55847fc2 (ceph): nuke: log what pid we are killing when we kill it
- 05:52 PM Revision 7586cde9 (ceph): qa/workunits/suites/pjd.sh: bash -x
- This will let us see what test is failing, exactly, and what its inputs
were. Hoping to help find #2187.
Signed-off... - 05:44 PM Revision 76efd977 (ceph): OSD: publish_map in init to initialize OSDService map
- Other areas rely on OSDService::get_map() to function, possibly before
activate_map is first called. In particular, ... - 05:25 PM Revision 675d6302 (ceph): ObjectCacher: fix cache_bytes_hit accounting
- Misses are not hits!
Signed-off-by: Josh Durgin <josh.durgin@inktank.com> - 02:35 PM Revision 4e1d973e (ceph): doc: Fixed heading text.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 02:35 PM Revision ebc57736 (ceph): doc: favicon.ico should be new Ceph icon.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 04:28 AM Revision 3a377c44 (ceph): doc: Overhauled Swift API documentation.
- Signed-off-by: John Wilkins <john.wilkins@inktank.com>
- 03:05 AM Revision 5c5ca4b8 (ceph): move cfuse + dbench from marginal to regression
- Fixed #1737, yay!
- 02:20 AM Revision aecf0031 (ceph): Merge branch 'next'
- 02:19 AM Revision d78235be (ceph): client: fix readdir locking
- Several of the readdir-related methods were not taking client_lock.
Fixes: #1737
Backport: argonaut
Signed-off-by: S... - 02:18 AM Revision 82a575c9 (ceph): client: fix leak of client_lock when not initialized
- Backport: argonaut
Signed-off-by: Sage Weil <sage@inktank.com>
07/17/2012
- 11:58 PM Revision 58e81c82 (ceph): OSD: handle_osd_ping: use service->get_osdmap()
- This way, we avoid grabbing the map_lock. Furthermore,
get curmap at the beginning of the method to ensure that
we s... - 11:58 PM Revision 90ddc5ae (ceph): OSD: use service.get_osdmap() in heartbeat(), don't grab map_lock
- service.get_osdmap() gives us sufficiently consist
access to the map state.
Signed-off-by: Samuel Just <sam.just@ink... - 11:51 PM Revision 32892c1e (ceph): doc/dev/osd_internals: add newlines before numbered lists
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 11:02 PM Revision fe4c658b (ceph): librados: simplify locking slightly
- No reason to hold mylock_all here.
Signed-off-by: Sage Weil <sage@inktank.com> - 07:40 PM Revision 199397dc (ceph): osd: default 'osd_preserve_trimmed_log = false'
- This option makes the osd skip zeroing old trimmed regions of the log. The
data is never read, since the xattrs indi... - 05:00 PM Revision 12dc0ad1 (ceph): ceph: archive mon data to a .tgz
- Saves bandwidth, time, and space.
- 04:54 PM Revision 24df8b1d (ceph): doc/dev: add osd_internals to toc
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 04:31 PM Revision 5a27f071 (ceph): doc/internals/osd_internals: fix indentation errors
- Signed-off-by: Samuel Just <sam.just@inktank.com>
- 03:36 PM Revision 36d0a355 (ceph): log: simplify log logic a bit
- Whether an entry is eligible to log/dump is independent of the channel it
is sent to. Some channels impose additiona... - 03:36 PM Revision 6490c84f (ceph): doc: discuss choice of pg_num
- Signed-off-by: Sage Weil <sage@inktank.com>
- 12:36 AM Revision abe05a3f (ceph): Merge branch 'next'
- 12:35 AM Revision 2869039b (ceph): Robustify ceph-rbdnamer and adapt udev rules
- Below is a patch which makes the ceph-rbdnamer script more robust and
fixes a problem with the rbd udev rules.
On ou... - 12:34 AM Revision 96587f39 (ceph): Robustify ceph-rbdnamer and adapt udev rules
- Below is a patch which makes the ceph-rbdnamer script more robust and
fixes a problem with the rbd udev rules.
On ou...
07/16/2012
- 11:30 PM Revision b0465496 (ceph): doc/radosgw/config.rst: mended small typo
- Signed-off-by: caleb miles <caleb.miles@inktank.com>
- 11:15 PM Revision 2a8c4db7 (ceph): Merge branch 'wip-mon-mkfs'
- Reviewed-by: Tommi Virtanen <tv@inktank.com>
- 11:14 PM Revision 54be9d09 (ceph): mon: stop doing rm -rf on mon mkfs
- Simply verify that the directory exists, or if it doesn't, create it.
Do nothing about its content.
Signed-off-by: S... - 11:14 PM Revision 6b1835a9 (ceph): vstart.sh: blow away mon directory on creation/start
- Now that ceph-mon doesn't blow away the mon data content, we need to.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:14 PM Revision 4e66a3b9 (ceph): mkcephfs: error out if mon data directory is not empty
- The ceph-mon --mkfs function no longer wipes out the directory; it is in
fact mostly a no-op that just verifies the d... - 11:14 PM Revision 4eec4fc5 (ceph): mkcephfs: nicer empty directory check
- From TV.
Signed-off-by: Sage Weil <sage@inktank.com> - 11:14 PM Revision 426384f6 (ceph): log: apply log_level to stderr/syslog logic
- In non-crash situations, we want to make sure the message is both below the
syslog/stderr threshold and also below th... - 11:13 PM Revision f9c1a6fb (ceph): Merge branch 'next'
- 11:02 PM Revision 52f96b9f (ceph): log: apply log_level to stderr/syslog logic
- In non-crash situations, we want to make sure the message is both below the
syslog/stderr threshold and also below th... - 10:53 PM Revision de524abd (ceph): log: dump logging levels in crash dump
- So you know what you are/are not seeing.
Signed-off-by: Sage Weil <sage@inktank.com> - 10:53 PM Revision d3c76f75 (ceph): Merge branch 'next'
- 10:43 PM Revision 3821f6c4 (ceph): PG: grab reference to pg in C_OSD_AppliedRecoveredObject
- Otherwise, accessing the pg via _applied_recovered_object
isn't safe. Using intrusive_ptr clarifies the reference
ow... - 10:37 PM Revision 8dafcc5c (ceph): log: fix event gather condition
- We should gather an event if it is below the log or gather threshold.
Previously we were only gathering if we were g... - 10:36 PM Revision 64f74500 (ceph): log: fix event gather condition
- We should gather an event if it is below the log or gather threshold.
Previously we were only gathering if we were g... - 09:18 PM Revision 4afa8925 (ceph): PG: use stats from primary after rewinding divergent entries
- If the osd recieving the info has divergent entries, it will
also have a "divergent" stat structure.
Probably fixes ... - 09:18 PM Revision d4410e4a (ceph): PG::RecoveryState::Stray::react(LogEvt&): set dirty_info/log
- We adjust the info and the log, so we must set dirty_info and
dirty_log to force writes.
Signed-off-by: Samuel Just ... - 09:18 PM Revision 5f602366 (ceph): Merge remote-tracking branch 'upstream/next'
- 09:11 PM Revision ec5cd6de (ceph): PG::RecoveryState::Stray::react(LogEvt&): reset last_pg_scrub
- We need to reset the last_pg_scrub data in the osd since we
are replacing the info.
Probably fixes #2453
In cases l... - 09:07 PM Revision c7fb964c (ceph): PG::RecoveryState::Stray::react(LogEvt&): reset last_pg_scrub
- We need to reset the last_pg_scrub data in the osd since we
are replacing the info.
Probably fixes #2453
In cases l... - 06:33 PM Revision 248cfadd (ceph): ReplicatedPG: don't warn if backfill peer stats don't match
- pinfo.stats might be wrong if we did log-based recovery on the
backfilled portion in addition to continuing backfill.... - 06:11 PM Revision 5d82a770 (ceph): doc/dev/osd_internals: OSD overview, pg removal, map/message handling
- This is a start on some osd internals documentation for new
developers.
Signed-off-by: Samuel Just <sam.just@inktank... - 05:59 PM Revision 470796b5 (ceph): CompatSet: users pass bit indices rather than masks
- CompatSet users number the Feature objects rather than
providing masks. Thus, we should do
mask |= (1 << f.id) rath... - 05:59 PM Revision 90381dc9 (ceph): OSD: set superblock compat_features on boot and mkfs
- Previously, we did not actually persist the osd compatibility
mask. Without persisting the current compat mask, a pr... - 05:59 PM Revision 12d70738 (ceph): PG: use write_info to set snap_collections in make_snap_collections
- At one point, snap_collections were written to a pg collection
attribute. Subsequently, they were moved to the bigin... - 05:59 PM Revision 1b8819bb (ceph): PG: Place info in biginfo object
- The purged_snaps set can grow without bound as snaps are
created and removed. Because the filestore doesn't
provide ... - 05:58 PM Revision bcb1073f (ceph): mon/MonitorStore: always O_TRUNC when writing states
- It is possible for a .new file to already exist, potentially with a
larger size. This would happen if:
- we were p... - 05:57 PM Revision 41a57077 (ceph): osd: based misdirected op role calc on acting set
- We want to look at the acting set here, nothing else. This was causing us
to erroneously queue ops for later (wastin... - 05:57 PM Revision b7814dbe (ceph): osd: based misdirected op role calc on acting set
- We want to look at the acting set here, nothing else. This was causing us
to erroneously queue ops for later (wastin... - 05:57 PM Revision 14d2efc4 (ceph): mon/MonitorStore: always O_TRUNC when writing states
- It is possible for a .new file to already exist, potentially with a
larger size. This would happen if:
- we were p... - 05:53 PM Revision ff0f4742 (ceph): set machine description to ctx.archive when auto-locking machines for a...
- 05:48 PM Revision e429da34 (ceph): Merge remote-tracking branch 'gh/bugfix-2022'
- Reviewed-by: Samuel Just <sam.just@inktank.com>
- 05:35 PM Revision 811665a8 (ceph): move cfuse + ffsb from marginal to regression
- This has had no failures.
- 04:41 PM Revision 88e2ad35 (ceph): move cfuse + fsx back into regression suite
- No failures in marginal. The objectcacher fixes that came out of the
rbd_fsx stuff probably fixed the original problem? - 04:12 PM Revision 47b38dd0 (ceph): Merge remote-tracking branch 'gh/bugfix-2779'
- Reviewed-by: Greg Farnum <greg@inktank.com>
- 05:03 AM Revision f94c7646 (ceph): mon: remove osds from [near]full sets when their stats are removed from...
- Greg points out that we could have a situation like:
- mon recovers..
- goes through osdmaps, notes an osd was rem... - 04:38 AM Revision fe576818 (ceph): mon/MonitorStore: always O_TRUNC when writing states
- It is possible for a .new file to already exist, potentially with a
larger size. This would happen if:
- we were p...
Also available in: Atom