Activity
From 06/15/2014 to 07/14/2014
07/14/2014
- 04:37 PM Bug #8817: enterprise-testing/1,2 RHEL 7.0 -- gui times out waiting for cluster.
- I'm gonna say this is conditionally something we can assume is working-ish until there's a hard horrid failure
- 01:25 PM Bug #8817: enterprise-testing/1,2 RHEL 7.0 -- gui times out waiting for cluster.
- I may have messed up setting up an OSD in the previous test (it came up running 2 out of 3
OSDS after the timeout). ... - 11:09 AM Bug #8817: enterprise-testing/1,2 RHEL 7.0 -- gui times out waiting for cluster.
- In the latest version (built 01:00 on 7/14), the No Cluster message occurred. However, clicking on the dashboard ico...
- 04:35 PM Feature #8779: calamari-ctl clear hangs, unkillably, if DB is open in cthulhu
- Not primary functionality; should verify that it's due to another database user, and see if there's anything calamari...
- 04:34 PM Bug #8791 (Resolved): logs are not visible
- Worked around in calamari commit a8ee07e781c90d24523adcd641d9ff88e52e1b7c. See also #8828.
- 04:29 PM Bug #8794 (Resolved): testing ice on ubuntu precise fails during ceph install
- mds-related packages reinserted (no change to build scripts; just stopped using local cache of ceph packages that had...
- 10:43 AM Fix #8828 (New): Remove log handling from cthulhu
- cthulhu is single threaded, and sending things like requests to fetch N lines of logs through it will make things chu...
- 07:26 AM Feature #8664 (In Progress): Add ability to get/set a crush map from the cluster
07/12/2014
- 12:44 PM Fix #8820 (Resolved): calamari: min_size is not set correctly when pool size is non-default
- Creating a pool with a non-default size will not set min_size properly.
We created a pool with a size of 2 with th...
07/11/2014
- 08:38 PM Bug #8817: enterprise-testing/1,2 RHEL 7.0 -- gui times out waiting for cluster.
- The calamari machine is vpm035. vpm030 is a ceph monitor, and osds are on vpm031 vpm032 and vpm033.
- 08:37 PM Bug #8817 (Closed): enterprise-testing/1,2 RHEL 7.0 -- gui times out waiting for cluster.
- Ceph is up. After hitting the ADD button to attach the ceph sites, the gui times out after 3 minutes.
It produces... - 08:22 PM Revision 2597aba8 (calamari): Merge pull request #159 from ceph/wip-fix-logs
- cthulhu: add log service function to avoid privilege probs calling salt directly from REST
Reviewed-by: Dan Mick <da... - 10:05 AM Revision 5cdfc887 (calamari): Detect SUSE Linux in get-flavor.sh
- Signed-off-by: Tim Serong <tserong@suse.com>
- 10:05 AM Revision 8f213d0b (calamari): Makefile: Install config files with mode 0644
- Unless a mode is explicitly specified when invoking `install`, it'll use
0755, which means we have config files insta... - 10:05 AM Revision c45c5f5b (calamari): .gitignore: Add *~, *.orig, *.rej
- Signed-off-by: Tim Serong <tserong@suse.com>
07/10/2014
- 08:35 PM Revision a8ee07e7 (calamari): cthulhu: absorb the salt from rest-api to get logs when the www use...
- Signed-off-by: Gregory Meno <gmeno@redhat.com>
- 11:45 AM Feature #8812 (New): RFE for calamari-ctl initialize: show Calamari URL/user
- Suggestion for the calamari installer, based on the Foreman installer, where it tells the user the URL and the userna...
- 07:12 AM Bug #8729: Building the trusty tarball image fails
- Dan Mick wrote:
> A community member has contributed: https://github.com/ceph/calamari/pull/153, or will soon, which... - 04:44 AM Revision f15abde8 (calamari): Untested code to chown -R <apacheuser> <salt-dirs>
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 12:38 AM Revision a6c692d6 (calamari): Merge pull request #157 from ceph/master_1.2_merge_gmeno
- Master 1.2 merge gmeno
07/09/2014
- 10:29 PM Bug #8794: testing ice on ubuntu precise fails during ceph install
- On reflection, while ceph-deploy maybe should change to allow this, I think in this case the problem is ICE's to solv...
- 04:54 PM Bug #8794: testing ice on ubuntu precise fails during ceph install
- This is ceph-deploy specifically asking for ceph-mds and ceph-fs-common, which means we can't remove them from the su...
- 04:34 PM Bug #8794 (Resolved): testing ice on ubuntu precise fails during ceph install
- Test machine: vpm115
tried to set up ceph cluster on vpm116
This happened while testing ice [https://download.ink... - 05:27 PM Bug #8720: Calamari: Pg status bar remains clean until refresh.
- I should catch up on email before I update issues.
- 04:48 PM Bug #8720: Calamari: Pg status bar remains clean until refresh.
- I have now seen this on RHel 6.5. Here's what I did.
I have 448/448 pg active and clean. The bar is all green
... - 11:53 AM Bug #8720: Calamari: Pg status bar remains clean until refresh.
- Steps to reproduce:
in one tab load dashboard
in another tab goto workbench take an osd out.
look at first tab w... - 02:06 PM Bug #8791 (Resolved): logs are not visible
- Logs show up as empty. The root cause is that calamari-web is attempting to call client.cmd() to invoke log_tail, bu...
- 03:28 AM Feature #8779: calamari-ctl clear hangs, unkillably, if DB is open in cthulhu
- Vaguely recall seeing this in development when something else was holding the database open, you don't by any chance ...
07/08/2014
- 09:44 PM Bug #8780 (Resolved): salt-minion didn't cleanly start ceph.py
- Okay. I might have had something not quite right in my setup.
I downbursted 5 new virtual machines, and vewwy vew... - 07:55 PM Bug #8780 (Need More Info): salt-minion didn't cleanly start ceph.py
- I looked at the monitor host; it had this in the minion log:...
- 07:17 PM Bug #8780 (Resolved): salt-minion didn't cleanly start ceph.py
- Running the centos 6.5 distro tarball (from https://download.inktank.com/enterprise-testing/1.2-rc4/)
Brought up C... - 05:52 PM Feature #8779: calamari-ctl clear hangs, unkillably, if DB is open in cthulhu
- kill with no arg (which IIRC is TERM) killed it, and it apparently had not cleared the db, as "initialize" did not re...
- 05:51 PM Feature #8779 (Rejected): calamari-ctl clear hangs, unkillably, if DB is open in cthulhu
- [ubuntu@vpm071 ice]$ sudo calamari-ctl clear --yes-i-am-sure
[INFO] Loading configuration..
[INFO] Dropping tables
... - 10:53 AM Bug #8768 (Duplicate): Rhel 7: Information displays a lot of unknown values.
07/07/2014
- 05:09 PM Bug #8720: Calamari: Pg status bar remains clean until refresh.
- Note: this problem seems to only happen when osds go down. Pg status is 0/xxx Active & Clean, but the status bar does...
- 04:27 PM Bug #8768 (Duplicate): Rhel 7: Information displays a lot of unknown values.
- John Wilkins came across this, and I verified that I see it on a cluster that I set up.
On the gui, run Manage(c... - 04:22 PM Revision d021c302 (calamari): calamari: fix lint error in TEMPLATE_DIRS setting
- Signed-off-by: Gregory Meno <gmeno@redhat.com>
- 04:03 PM Revision 558eb247 (calamari): fix travis looking for requirements in the wrong place
- Signed-off-by: Gregory Meno <gmeno@redhat.com>
- 03:32 PM Bug #8729: Building the trusty tarball image fails
- A community member has contributed: https://github.com/ceph/calamari/pull/153, or will soon, which should in theory h...
- 03:19 PM Cleanup #8767 (Closed): remove salt/systemd service workarounds
- Several places in the .sls files dealing with services (in salt/local) are clunky because of salt bug https://github....
- 11:05 AM Revision 7ebf7fac (calamari): Merge remote-tracking branch 'remotes/origin/1.2'
- Conflicts:
README.rst
doc/development/building_packages.rst - 07:28 AM Feature #8755 (Duplicate): make calamari-ctl generate postgres passwords in config readable only ...
- 06:45 AM Feature #8755: make calamari-ctl generate postgres passwords in config readable only to www user
- This is a dupe of #8064
- 06:16 AM Feature #8755 (Duplicate): make calamari-ctl generate postgres passwords in config readable only ...
- In the calamari-server rpm, in calamari/calamari.conf:
db_path = postgresql://calamari:27HbZwr*g@localhost/calamar... - 06:18 AM Feature #8756 (New): Update django to fix known security flaws
- https://www.djangoproject.com/weblog/2014/apr/21/security/
07/06/2014
- 03:44 AM Revision 86429258 (calamari-clients): Clean up salt-bootstrap: just use urllib
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 02:29 AM Revision b2d3538b (calamari): Missed edit of path for bootstrap
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 01:37 AM Revision 42975cdf (calamari): Oops, nope, rhel7 not quite done yet
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
07/05/2014
- 10:46 PM Revision af54b339 (calamari): Clean up salt-bootstrap: 1) stable works for rhel7 2) use urllib ev...
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
07/03/2014
- 10:00 PM Bug #8720: Calamari: Pg status bar remains clean until refresh.
- Note: This happens on both Chrome and Firefox
- 09:11 PM Bug #8731 (Won't Fix): Fix SELinux for Calamari on RHEL7
- Decision for now is to document "disable SELinux". Reopen if we want to revisit this.
- 09:09 PM Bug #8721: Calamari on RHEL does not come up after reboot.
- Fixed in 1.2 commit 1a7f62a06daa2864d2c36fb022c0d5533fe7bb26
- 09:08 PM Bug #8721 (Resolved): Calamari on RHEL does not come up after reboot.
- 09:06 PM Bug #8742: calamari build: some failures aren't reported by vagrant ssh
- The build had a failure midway through while trying to make el6 in repobuild which was not transmitted out to a failu...
- 09:06 PM Bug #8742: calamari build: some failures aren't reported by vagrant ssh
- The build had a failure midway through while trying to make el6 in repobuild which was not transmitted out to a failu...
- 05:29 PM Bug #8742 (Resolved): calamari build: some failures aren't reported by vagrant ssh
- Running
./ice_repo_tgz.py -v -b v0.81 -o centos6.5
Resulted in:... - 08:11 PM Bug #8448 (Resolved): ICE: calamari-ctl initialize error when not run as root.
- Looks good.
- 11:37 AM Bug #8448 (Fix Under Review): ICE: calamari-ctl initialize error when not run as root.
- 07:03 AM Bug #8448: ICE: calamari-ctl initialize error when not run as root.
- The way that I think I'll handle part 1 is:
add a cmdline switch --devmode
the behavior is then:
if not args.de... - 06:39 AM Bug #8448 (In Progress): ICE: calamari-ctl initialize error when not run as root.
- 07:38 PM Revision 7257c30b (calamari): Merge pull request #156 from ceph/wip-8448-1.2
- cthulhu: add flag to signal that we do not need root privs
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 06:55 PM Revision b3b14d8d (calamari): calamari-ctl: add flag to signal that we do not need root privs
- Signed-off-by: Gregory Meno <gmeno@redhat.com>
- 06:47 PM Revision 2b765cc0 (calamari): Merge pull request #155 from ceph/wip-8721
- Add salt/calamari-ctl machinery to start and enable services
- 06:39 PM Revision 1a7f62a0 (calamari): Add salt/calamari-ctl machinery to start and enable services
- Work around salt systemd service bug for RHEL7
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 06:26 PM Bug #8743: Salt grains empty when server is in accepted state, rev 2
- (The workaround is to add r-x permissions to those dirs. It's not clear if this is any sort of security hole.)
- 06:25 PM Bug #8743 (Resolved): Salt grains empty when server is in accepted state, rev 2
- Start by tracking this down in salt.
#7783 documented a problem with permissions accessing grains for the minion; ...
07/02/2014
- 11:39 PM Revision ea152605 (calamari): Merge pull request #154 from ceph/wip-8723
- pycairo is a dependency of graphite-web (masked by extra packages on vps...
- 11:14 PM Revision 2b4ced2c (calamari): pycairo is a dependency of graphite-web (masked by extra packages o...
- Fixes: #8723
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 10:12 PM Bug #8729: Building the trusty tarball image fails
- There is no calamari build for trusty yet either, so it's not like the tarball creation would get very far.
- 12:54 PM Bug #8729 (Closed): Building the trusty tarball image fails
- This is the last few lines of the ./ice_repo_tgz.py -v -b v0.81 -o trusty
command output. I could provide more if n... - 10:11 PM Bug #8350 (Resolved): UI: IOPs count frequently displays 0 incorrectly
- The fix for this has been merged.
- 10:07 PM Bug #8548 (Resolved): UI: IOPS do not appear to be detected.
- 05:40 PM Bug #8548: UI: IOPS do not appear to be detected.
- No, build was made with branch that didn't contain fix fro this bug
- 04:18 PM Bug #8548: UI: IOPS do not appear to be detected.
- Diamond does not seem to come up in 6.4. Is this expected right now?
- 10:05 PM Fix #7783: Salt grains empty when server is in accepted state.
- It turns out there are some cases where access also requires root privs (it seems to be the case on RHEL7 that severa...
- 09:59 PM Bug #8448: ICE: calamari-ctl initialize error when not run as root.
- part 2:
--- a/ice_setup.py
+++ b/ice_setup.py
@@ -1242,7 +1242,7 @@ def default():
ceph_gpg_url,
... - 09:57 PM Bug #8721 (Fix Under Review): Calamari on RHEL does not come up after reboot.
- wip-8721 seems to work on rhel7.
- 04:44 PM Bug #8721: Calamari on RHEL does not come up after reboot.
- httpd and salt-master are both restarted by calamari-server's postinstall; supervisord is restarted by calamari-ctl i...
- 05:30 PM Bug #8731 (New): Fix SELinux for Calamari on RHEL7
- Found several SELinux errors in Calamari running on RHEL7
For example:
/var/log/audit/audit.log
type=AVC msg=aud... - 02:52 PM Bug #8596: ICE: ice_setup.py fails on Centos 6.3
- Yeah, I see the issue now; apparently 6.3 included a package "redhat-lsb" and later included "redhat-lsb-core", and i...
- 01:34 PM Bug #8596: ICE: ice_setup.py fails on Centos 6.3
- Rechecked. This is still happening
07/01/2014
- 09:45 PM Bug #8721: Calamari on RHEL does not come up after reboot.
- (because the same is true for salt-master.)
- 09:44 PM Bug #8721: Calamari on RHEL does not come up after reboot.
- So, from a discussion on #systemd, I am now under the impression that Fedora et. al. expect part of the app installat...
- 09:31 PM Bug #8721: Calamari on RHEL does not come up after reboot.
- Note: This also happens on Centos 6.5
- 07:41 PM Bug #8721 (In Progress): Calamari on RHEL does not come up after reboot.
- 07:41 PM Bug #8721: Calamari on RHEL does not come up after reboot.
- The same is true of supervisord. I guess this is a difference between the packages on RHEL7 vs RHEL<7, and I guess a...
- 07:34 PM Bug #8721: Calamari on RHEL does not come up after reboot.
- 07:34 PM Bug #8721: Calamari on RHEL does not come up after reboot.
- So I note that chkconfig httpd says
Note: Forwarding request to 'systemctl is-enabled httpd.service'.
disabled
... - 06:34 PM Bug #8721 (Resolved): Calamari on RHEL does not come up after reboot.
- If you reboot the Calamari server, the UI cannot connect back.
The problem is probably due to all sorts of service... - 09:16 PM Bug #8720: Calamari: Pg status bar remains clean until refresh.
- This seems to be a problem on RHEL 7. Centos 6.5 works much better. The Bar changes to yellow and back to green wit...
- 05:21 PM Bug #8720: Calamari: Pg status bar remains clean until refresh.
- Note. When I brought the OSD back up, and PG/Status read 512/512 Active & Clean, I needed to hit the refresh to get ...
- 05:15 PM Bug #8720 (Resolved): Calamari: Pg status bar remains clean until refresh.
- RHEL 7 -- 1 Calamari Server, 1 Monitor Server, 3 OSDs.
OSD was marked down via the UI. 2 of 3 osds are up accordi... - 08:56 PM Bug #8688 (Resolved): ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- Appears to be fixed.
- 02:05 PM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- Verified all the distros are pulling salt 2014.1.5 now.
- 01:03 PM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- We've been on an "update the iceball with each bug fix" schedule for some time; I don't know that version will help.
... - 08:11 AM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- Warren:
Would you please provide more info?
What version/md5sum of iceball were you using?
Did vpm005 get setu... - 08:07 AM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- FIXED:
[ubuntu@vpm004 ~]$ sudo salt '*' saltutil.sync_modules
vpm005.front.sepia.ceph.com:
- modules.ceph
... - 08:07 AM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- [ubuntu@vpm005 ~]$ sudo yum install salt
Loaded plugins: fastestmirror
Repository 'calamari-minion' is missing name... - 08:05 AM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- checking versions of salt indicates a small mismatch
MASTER:
[ubuntu@vpm004 ~]$ rpm -qa | grep salt
salt-master-... - 07:36 AM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- Things I've tried so far:
looking at cthulhu.log
rejecting salt keys of each node in the cluster not running a moni... - 07:27 AM Bug #8688 (Need More Info): ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- 2014-06-28 01:49:09,832 [salt.minion ][WARNING ] TypeError encountered executing salt...
- 06:42 PM Bug #8548 (Resolved): UI: IOPS do not appear to be detected.
- Appears to be working now.
- 04:55 PM Bug #8548: UI: IOPS do not appear to be detected.
- 08:49 AM Bug #8548: UI: IOPS do not appear to be detected.
- These are the same messages that https://github.com/ceph/calamari/commit/4cdcba721bc84791e44fec0f1a0af49101cadd08
fi... - 04:54 PM Bug #8451 (Can't reproduce): ICE: Wheezy Hash sum errors in ice_setup.py
- Assuming this is gone/was a network hiccup.
- 01:13 PM Bug #8525 (Closed): ICE: ceph_deploy: ceph --version fails
- 01:13 PM Bug #8525: ICE: ceph_deploy: ceph --version fails
- Pretty sure the dependencies have been fixed now. If you still see this, Warren, please reopen.
- 01:05 PM Bug #8448: ICE: calamari-ctl initialize error when not run as root.
- yeah - bear in mind that sudo is only for ubuntu et al, so we shouldn't tell people to use it when they're on RHEL. ...
- 11:52 AM Bug #8448: ICE: calamari-ctl initialize error when not run as root.
- The work here is to:
1. notice when calamari-ctl needs elevated privs. Catch the error and suggest that the user r... - 10:25 AM Bug #8468 (Resolved): Change configuration so that tests against may be run against an external c...
- 05:08 AM Revision 4cdcba72 (calamari): Merge pull request #152 from ceph/wip-fix-diamond-requisites
- fix files to properly require the package
Believed to fix 8548
References: #8548
Reviewed-by: Dan Mick <dan.mick@ink...
06/30/2014
- 11:34 PM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- I don't have a good idea what to look for here. Restarted cthulhu in debug mode; poked a bit at the postgres db; don...
- 10:59 AM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- I left the configuration up. When I brought up the gui again, I get the message:...
- 10:54 PM Bug #8548: UI: IOPS do not appear to be detected.
- Indeed diamond isn't running; there are these warnings in /var/log/salt/minion:
2014-06-30 20:14:22,489 [salt.stat... - 05:34 PM Bug #8548: UI: IOPS do not appear to be detected.
- vpm050 locked wusui@aardvark "Rhel 7 Server"
vpm063 locked wusui@aardvark "Rhel 7 Ceph Monitor"
vpm072 locked... - 05:28 PM Bug #8548: UI: IOPS do not appear to be detected.
- Note: Greg. I assigned this to you because I heard that you were working on some diamond startup issues.
- 05:25 PM Bug #8548 (In Progress): UI: IOPS do not appear to be detected.
- See the above comments. It looks like Diamond is not starting correctly after the page add prompt is hit.
- 05:24 PM Bug #8548: UI: IOPS do not appear to be detected.
- Further details:
I reinstalled RHEL 7 from scratch on a 5 node cluster.
I finished running ceph-deploy calama... - 05:20 PM Bug #8548: UI: IOPS do not appear to be detected.
- Further details:
I reinstalled RHEL 7 from scratch on a 5 node cluster.
I finished running ceph-deploy calama... - 02:00 PM Bug #8548: UI: IOPS do not appear to be detected.
- Latest version -- Diamond does not appear to be running. I brought up 1 Calamari Server, 1 mon server, and 3 osd ser...
- 08:52 PM Bug #8451: ICE: Wheezy Hash sum errors in ice_setup.py
- So the first error appears to have been a problem communicating with the Ubuntu package mirror or an incomplete/damag...
- 06:26 PM Bug #8597 (Closed): Calamari: message not clear, or my brain is not.
- Messages have been fixed.
- 06:18 PM Bug #8552 (Closed): Unusual behavior after rebooting Calamari.
- Does not appear to be happening anymore.
- 03:02 PM Revision b150d715 (calamari): Merge pull request #148 from ceph/wip-external-tests-docs
- docs: adds instructions for running tests on external calamari
- 03:02 PM Revision e02c90ab (calamari): Merge pull request #143 from ceph/wip-external-tests
- Wip external tests
- 03:01 PM Revision b3a40b20 (calamari): Merge pull request #147 from ceph/wip-sim
- minion_sim: stats path updates
- 01:03 PM Bug #8689 (Can't reproduce): ICE: Precise: Blank screen on gui
- Okay. I tried this twice today and both times, the gui came up okay. I'm going to close this for now.
06/29/2014
- 09:19 AM Bug #8689: ICE: Precise: Blank screen on gui
- For GUI issues like this it is always useful to check the javascript console for errors (right click, Inspect Element...
06/27/2014
- 08:31 PM Bug #8689 (Can't reproduce): ICE: Precise: Blank screen on gui
- I brought up Precise and when I logged in to the gui, all I got is a grey screen.
I am not sure how fragile the in... - 07:48 PM Revision 17d0387c (calamari): fix files to properly require the package
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 07:17 PM Bug #8649 (Resolved): rHEL 7 -- calamari-ctl initialize failed.
- Looks good.
- 06:57 PM Bug #8688: ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- Dunno if you are the right person for this, but I am leaving vms up for someone to look at so I thought that it would...
- 06:56 PM Bug #8688 (Resolved): ICE: Centos 6.5 -- no EPEL -- Ceph is up but cannot be found by gui.
- One monitor, 3 OSDs. All four are shown in the gui, but after the add button is hit, the three minute timeout limit ...
- 01:42 AM Revision 378d76a6 (calamari): Merge pull request #150 from ceph/wip-fix-diamond-ubuntu
- salt: fix conditional inclusion of file: diamond-init-config
Reviewed-by: Dan Mick <dan.mick@inktank.com> - 01:17 AM Revision 804782e4 (calamari): salt: fix conditional inclusion of file: diamond-init-config
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
06/26/2014
- 02:59 AM Revision a2a22bc8 (calamari): syntax error in prior commit
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
06/25/2014
- 10:45 PM Revision b646f6c4 (calamari): Merge pull request #149 from ceph/wip-8669
- graphite template dir needs to vary based on python minor version
- 10:31 PM Revision e69840f7 (calamari): graphite template dir needs to vary based on python minor version
- Fixes: #8669
Signed-off-by: Dan Mick <dan.mick@inktank.com> - 10:26 PM Bug #8615 (Duplicate): ICE: ceph-deploy calamari connect with more than one minion missing baseurl
- #8671 documents this again, but I've figured it out I think.
- 10:23 PM Bug #8607 (Can't reproduce): ICE: rhel7 ice_setup.py failure
- Don't know what was wrong, but it's gone.
- 07:08 PM Bug #8649: rHEL 7 -- calamari-ctl initialize failed.
- So for future reference:
1) the /tmp/2014-whatever is just 'some error occurred', and there's a *lot* of errors th... - 06:59 PM Bug #8649: rHEL 7 -- calamari-ctl initialize failed.
- #!@@!!Z -- I forgot to disable EPEL... I will look at this further.
- 06:49 PM Bug #8649: rHEL 7 -- calamari-ctl initialize failed.
- Maybe I'm doing something wrong. I just got that argparse problem again on this 6.5 version too.
- 06:47 PM Bug #8649 (In Progress): rHEL 7 -- calamari-ctl initialize failed.
- I just had this pop up again on Centos 6.5. See /tmp/2014-06-26_0544.txt on vpm003. I'll leave it up.
- 06:18 PM Bug #8672 (Closed): Manage(ceph) > OSD shows only one host.
- 06:18 PM Bug #8672: Manage(ceph) > OSD shows only one host.
- Okay. I only have one osd configured. Pilot error.
- 04:28 PM Bug #8672 (Closed): Manage(ceph) > OSD shows only one host.
- I brought up calamari with two ceph hosts. I saw them both listed after I connected them, and I see them on the Mana...
- 05:25 PM Bug #8669 (Resolved): graphite is broken
- 03:19 PM Bug #8669: graphite is broken
- Apparently this has always been broken:
$ grep 2.7 calamari-web/calamari_web/settings.py
TEMPLATE_DIRS = (os.path... - 03:12 PM Bug #8669: graphite is broken
- django template directory search is broken; looking for /opt/calamari/venv/lib/python2.7/site-packages/graphite/templ...
- 02:44 PM Bug #8669 (Resolved): graphite is broken
- In current setup, graphite appears broken; graphite/dashboard gives a 500. I don't understand how GRAPHITE_API_PREFI...
- 03:11 PM Revision a0b68896 (calamari): tests: remove port number
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision 11290baf (calamari): cthulhu: fix PEP8 warning
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision c3d6c51d (calamari): tests: stop iterating the characters in the log path
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision b1fb1682 (calamari): tests: skip changing the crush rules set of a pool if there is only...
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision 053037f7 (calamari): tests: skip external tests that require calamari resets
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision e9053e90 (calamari): tests: improves perfomance of inter-test reset for live clusters
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision c5081c42 (calamari): tests: stop setting osds in, they should alredy be in, stop tests i...
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision 67df3b5f (calamari): tests: choose which bootstrap to use based on setting in conf
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision e07a57f3 (calamari): tests: stop passing JSON around. working bootstrap
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision 22c83ec2 (calamari): tests: change get_timeout_scaling_factor to return int
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision 2b640e27 (calamari): tests: Move api location and creds into test config template
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:11 PM Revision 675427d3 (calamari): tests: Reduce operations on pools during inter-test reset
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 03:10 PM Revision 9a7ff5fc (calamari): docs: adds instructions for running tests on external calamari
- Signed-off-by: Gregory Meno <gregory.meno@inktank.com>
- 10:50 AM Feature #8667 (Resolved): CRUSH: override 'crush location' in the ceph config so that modificatio...
- TBD
- 10:49 AM Feature #8666 (Resolved): resolve an OSD to crush node, decorate the OSD details with that
- /api/v2/cluster/<FSID>/osd
GET -> OsdViewSet.list(request, fsid)
/api/v2/cluster/<FSID>/osd/<osd_id>
GET -... - 10:48 AM Feature #8665 (Resolved): Add CRUD to calamari api for crush nodes
- /api/vs/cluster/<FSID>/crush_node
GET -> CrushNodeViewSet.list(request, fsid)
POST -> CrushNodeViewSet.creat... - 10:43 AM Feature #8664 (Resolved): Add ability to get/set a crush map from the cluster
- /api/vs/cluster/<FSID>/crush_map
GET -> CrushMapViewSet.retreive(request, fsid)
POST -> CrushMapViewSet.repl... - 10:41 AM Feature #8663 (New): CRUSH: Add CRUD to calamari api for crush_rules
- /api/v2/cluster/<FSID>/crush_rule/<rule_id>
GET -> CrushRuleViewSet.retreive(request, fsid, rule_id)
POST ... - 10:35 AM Bug #8662 (Resolved): apt-get update in the bootstrap scrip doesn't exit cleanly
- Failed to fetch http://vpm026.front.sepia.ceph.com/static/ubuntu/dists/precise/main/binary-i386/Packages 404 Not Fo...
- 10:13 AM Bug #8524 (Resolved): ICE: cthulhu uses logging.logger.getChild, must not (not in 2.6)
06/24/2014
- 11:29 PM Bug #8658 (Resolved): ICE: ice_repo_tgz.py fails on rhel7:
- Yes, we discussed this, fixed in wip-rhel7 branch of ice-tools, commit 637e540f41f439bc90ffd01ce5f58543f66db938
- 04:47 PM Bug #8658 (Resolved): ICE: ice_repo_tgz.py fails on rhel7:
- ./ice_repo_tgz.py -v -b v0.80.1 -o rhel7
Evetually displays:... - 08:50 PM Revision afa0db65 (calamari): rest-api: workaround for salt grain permissions
- /var/cache/salt used to be readable from apache, where
calamari_rest used it to read cached minion grains. That
does... - 07:09 PM Bug #8627: ice-setup writes ceph.repo file with no 'name' key; causes yum warnings
- Sadly, not quite end-to-end: custom_repo also does not write the 'name' key:...
- 06:00 PM Revision de7a13fa (calamari): Merge pull request #145 from zhurongze/fix-getChild-AttributeError-...
- Implement logging.getChild() with join() (not in 2.6) in other places
- 05:42 PM Revision 56833422 (calamari): Merge pull request #146 from zhurongze/fix-socket-path-pattern-match
- Fix socket path pattern match
- 02:07 PM Feature #8657 (Resolved): Epic: Crush Management
- 09:42 AM Bug #8615: ICE: ceph-deploy calamari connect with more than one minion missing baseurl
- tarball construction gets the latest ice-setup.py. My guess is that we just need to retest and verify this is gone.
- 07:00 AM Bug #8615: ICE: ceph-deploy calamari connect with more than one minion missing baseurl
- This looks like an old version of ice_setup.py
Warren can you paste the cephdeploy.conf file to see the contents?
... - 09:21 AM Bug #8649 (Resolved): rHEL 7 -- calamari-ctl initialize failed.
- 05:34 AM Revision 3ef7bba4 (calamari): Put centos-bootstrap-salt.sh in correct directory
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 05:00 AM Revision 38234e03 (calamari): curl now failing on centos 6.4; use 'urllib' workaround with stable...
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 03:57 AM Revision 65062242 (calamari): Mismerge of postgres.sls left extra {% endif %}
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 02:51 AM Revision ce37f47a (calamari): salt: Fix socket path pattern match
- The regular expression "^(.*)-(.*)\.(.*).asok$" for socket path can't catch
the example "ceph-mon.www-2.abc.com.asok"... - 02:49 AM Revision 35f65321 (calamari): Implement logging.getChild() with join() (not in 2.6) in other places
- the patch is like 685a57ed762a9a6e7aea420a6189c3651b03f018
Signed-off-by: Rongze Zhu <zrzhit@gmail.com> - 12:13 AM Bug #8650 (Resolved): centos tarball missing python-jinja2-26
- 743ae6f05bf5cff516964254ee7db8f259f5e0ca in ice-tools adds back python-jinja2-26
06/23/2014
- 11:53 PM Bug #8650 (Resolved): centos tarball missing python-jinja2-26
- This was masked by the underlying cause of #8525, test machines with EPEL configured
- 11:39 PM Bug #8525: ICE: ceph_deploy: ceph --version fails
- ...and apparently exposes another bug, python-flask requires python-jinja2-26. sigh. See #8650
- 11:31 PM Bug #8525: ICE: ceph_deploy: ceph --version fails
- I think this is happening because the VMs have EPEL defined, and the EPEL version of ceph v0.80.1 is incompatible. T...
- 11:00 PM Bug #8615: ICE: ceph-deploy calamari connect with more than one minion missing baseurl
- Alfredo, can you investigate please?
- 10:58 PM Bug #8649: rHEL 7 -- calamari-ctl initialize failed.
- This was a merge error that left an extra "endif" in the postgres.sls file. Fixed with
65062242c78695e616e1b634077a... - 06:43 PM Bug #8649: rHEL 7 -- calamari-ctl initialize failed.
- Note. this problem also occurs on precise.
- 02:38 PM Bug #8649 (Resolved): rHEL 7 -- calamari-ctl initialize failed.
Got the following message:...- 02:00 PM Bug #8626 (Resolved): gperftools empty
- 9fac9aca872e58fb011fd734c3a78e714c76f89f in ice-tools
- 01:59 PM Bug #8618 (Resolved): rhel7: apache permission model changes in 2.4
- 319fd6afc4f12ba8d8bc2129fe699d736f33be3f in 1.2 branch
- 01:59 PM Bug #8617 (Resolved): rhel7: salt service provider is broken
- 2ada825e62ebcba6144a488813ffa51ae14365fe in 1.2 branch
- 01:17 PM Revision 20fdbca0 (calamari): minion_sim: stats path updates
- Signed-off-by: John Spray <john.spray@redhat.com>
06/20/2014
- 03:54 PM Revision df2dcdc2 (calamari-clients): Merge pull request #36 from ceph/1-2-issue-8597
- 1.2 - #8597
- 02:57 PM Revision 65e801ed (calamari): Merge branch 'wip-rhel7' into 1.2
- Conflicts:
salt/local/postgres.sls - 06:02 AM Bug #8627 (Resolved): ice-setup writes ceph.repo file with no 'name' key; causes yum warnings
- Fixed with commit 0fcc8f1 on the ice-setup repository, master branch.
06/19/2014
- 06:52 PM Revision 319fd6af (calamari): Add a specialization of httpd conf for RHEL7/Apache 2.4's new auth ...
- Also, make get-flavor.sh a bit more careful, and call 'rh' 'el6' to reflect
the fact that it's been used for both Cen... - 10:41 AM Bug #8627: ice-setup writes ceph.repo file with no 'name' key; causes yum warnings
- seems also to affect the calamari-minions repo.
- 08:53 AM Bug #8627 (Resolved): ice-setup writes ceph.repo file with no 'name' key; causes yum warnings
- It's a nit, but it's a warning on every yum operation that we can easily remove.
- 07:32 AM Bug #8626 (Resolved): gperftools empty
- The gperftools package is now in EPEL (for RHEL7); however, it's a metapackage, and the real content is in gperftools...
06/18/2014
- 12:01 PM Bug #8528 (Resolved): ICE: Centos ceph-deploy mon create failing
- The resolution is: use leveldb 1.7 for now on RPM hosts.
- 11:59 AM Bug #8619 (Can't reproduce): rhel7: crashes on access
- 11:59 AM Bug #8619: rhel7: crashes on access
- yes, with enough memory, the service comes up
- 10:51 AM Bug #8619: rhel7: crashes on access
- A retest showed partial slow success, and it occurred to me that my test vm was provisioned with only 512M; trying ag...
06/17/2014
- 02:16 AM Revision 2ada825e (calamari): Workarounds for salt bug https://github.com/saltstack/salt/pull/12316
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
06/16/2014
- 09:49 PM Revision f0989c5e (calamari): HACK: paths for cthulhu+vstart
- 09:48 PM Revision 3355cf43 (calamari): mon_remote: first cut, almost works with vstart
- 09:48 PM Revision 84d59089 (calamari): remote/base: clear up
- 08:33 PM Bug #8619: rhel7: crashes on access
- 08:33 PM Bug #8619: rhel7: crashes on access
- As yet undiagnosed crash:...
- 08:31 PM Bug #8619 (Can't reproduce): rhel7: crashes on access
- 08:30 PM Bug #8618 (Resolved): rhel7: apache permission model changes in 2.4
- apache throws permission errors: see http://dabase.com/blog/AH01630:_client_denied_by_server_configuration/
Must n... - 08:28 PM Bug #8617 (Resolved): rhel7: salt service provider is broken
- See https://github.com/saltstack/salt/pull/12316 . The service provider needs to be selecting systemd, but is not.
... - 07:51 PM Bug #8525: ICE: ceph_deploy: ceph --version fails
- This still happens on RHEL 6.4. probably on others as well.
- 06:45 PM Bug #8615: ICE: ceph-deploy calamari connect with more than one minion missing baseurl
- This was RHEL 6.4 The ceph browser page also reports only one host reporting
- 06:45 PM Bug #8615 (Duplicate): ICE: ceph-deploy calamari connect with more than one minion missing baseurl
- ceph-deploy calamari connect vpm091 vpm098 displayed the following:...
- 06:32 PM Revision f5eeea00 (calamari): Workarounds for salt bug https://github.com/saltstack/salt/pull/12316
- Signed-off-by: Dan Mick <dan.mick@inktank.com>
- 05:03 PM Revision 6d38a6ad (calamari-clients): Issue #8597: do not append 'completed'.
- Because the messages are generated server side this often creates
confusing sentence construction. Remove all text fr... - 04:58 PM Revision 3e3686da (calamari-clients): Issue #8597: do not append 'completed'.
- Because the messages are generated server side this often creates
confusing sentence construction. Remove all text fr... - 01:57 PM Revision 7dc30ebe (calamari): fixup for salt_remote
- 10:36 AM Bug #8607 (Can't reproduce): ICE: rhel7 ice_setup.py failure
- yum install calamari-clients fails:...
- 10:24 AM Feature #6898: Per-pool PG details and state counters
- NB there is currently not a comprehensive list of PG states maintained in the calamari backend (to sidestep scaling i...
- 10:23 AM Feature #6946: investigate firewall config for Diamond agents
- I'm not sure we (Calamari) can really do much better than opening the ports we need (and we are using graphite's "saf...
- 09:59 AM Bug #8597: Calamari: message not clear, or my brain is not.
- calamari-clients master branch commit 3e3686daef55255c9ba29b5e8d4ab69cb421c388
Issue #8597: do not append 'com... - 05:59 AM Bug #8597: Calamari: message not clear, or my brain is not.
- I agree that this message should be clearer. It's technically accurate, in that the task of initiating the scrub has...
- 03:45 AM Feature #6855 (Rejected): Package Calamari variant of Salt minion
- I don't think we would want to do this any more. Now that Calamari is open source, the preferred approach to deal wi...
Also available in: Atom