[Bug 1432644] Re: VM permanently tries to read /dev/shm/lttng-ust-wait-5

2015-04-02 Thread Sage Weil
FWIW, we are disabling the lttng support in the final hammer release to avoid this issue (until we come up with a better solution). -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in Ubuntu. https://bugs.launchpad.net/bugs/1432644

Re: [Bug 1400299] Re: rocksdb ftbfs on ppc64el with -03

2014-12-08 Thread Sage Weil
On Mon, 8 Dec 2014, Ken Dreyer wrote: We should get this reported to Ceph upstream. Ceph is bundling an old copy of rocksdb (ugh), and I wonder if this is fixed in a later version of rocksdb. By the way, my hunch is that the reason that Ceph bundles rocksdb is because it was not packaged

[Bug 1278466] Re: [FFe] ceph firefly stable release

2014-03-18 Thread Sage Weil
Dave: correct. There are a few very minor changes going into 0.79 to make the final CLI/REST API experience good. We expect nothing but performance and bug fixes for 0.80. -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in

[Bug 1261501] Re: ceph osds unbalanced

2013-12-16 Thread Sage Weil
This may also be due to legacy CRUSH behavior that is fixed in bobtail (though not made default until recently). In newer Ceph versions, you can do 'ceph osd crush tunables optimal' to get the improved mapping (expect some data movement). I suggest you upgrade, in any case! -- You received

Re: [Bug 1227370] Re: ceph osd weight get reset to 1 on daemon restart

2013-09-18 Thread Sage Weil
This is fixed/changed in newer versions. The current releases will weight based on df reported size. Or, weighting on start can simply be diasabled... That is probably the thing to do for 0.48. James Troup james.tr...@canonical.com wrote: The upstart job is shipped by the charm, not in the

[Bug 1215014] Re: ceph-osd from Folsom leaks memory constantly

2013-08-21 Thread Sage Weil
There are known leaks in cuttlefish and bobtail, but we have not backported fixes because they are very very slow. Dumpling (0.67.x) is leak free. -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in Ubuntu.

[Bug 1129217] Re: rbd showmapped has trailing whitespace which break rbd monitor

2013-02-18 Thread Sage Weil
What about just iterating over the device nodes that udev creates? /dev/rbd/$pool/$image -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in Ubuntu. https://bugs.launchpad.net/bugs/1129217 Title: rbd showmapped has trailing

[Bug 1098320] Re: ceph: default crush rule does not suit multi-OSD deployments

2013-01-21 Thread Sage Weil
Is it important that this is backported to argonaut, given bobtail will be in 13.04? -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in Ubuntu. https://bugs.launchpad.net/bugs/1098320 Title: ceph: default crush rule does not

[Bug 1099900] Re: rados parsing error with hostnames in mon_host

2013-01-15 Thread Sage Weil
I take that back: this is already fixed in v0.48.3argonaut, which you should be upgrading to anyway due to some critical bug fixes. -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in Ubuntu. https://bugs.launchpad.net/bugs/1099900

[Bug 1099900] Re: rados parsing error with hostnames in mon_host

2013-01-15 Thread Sage Weil
s/case/code/ -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in Ubuntu. https://bugs.launchpad.net/bugs/1099900 Title: rados parsing error with hostnames in mon_host To manage notifications about this bug go to:

[Bug 1099900] Re: rados parsing error with hostnames in mon_host

2013-01-15 Thread Sage Weil
This is fixed in the mainline and bobtail, but not in argonaut. In newer case, any of [,; ] work. -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in Ubuntu. https://bugs.launchpad.net/bugs/1099900 Title: rados parsing error

[Bug 1099900] Re: rados parsing error with hostnames in mon_host

2013-01-15 Thread Sage Weil
commit 1c382869ba6aff0c159103a8267045c6eb600114 in mainline, commit ea42eb1f7866fed0e843257ce05e34503fd0129c in teh argonaut stable branch. I just did a quick test and it appears to be okay on master. Can you confirm that there were problems on anything other than v0.48.2? -- You received this

[Bug 1079941] Re: -v version output is broken

2012-11-17 Thread Sage Weil
It's not that the file gets stripped out, but that it is generated by 'make dist' when the tarballs are built. There are --with-system-libs3 and --with-system-leveldb configure flags that ignore the bundled library code. Would it help if those were the defaults? Or is there a fundamental issue

[Bug 1056462] Re: The '--print' option to ceph-authtool is incorrect

2012-10-08 Thread Sage Weil
I've also queued this patch in the stable branch so that it'll be included in the next stable point release (0.48.3). -- You received this bug notification because you are a member of Ubuntu Server Team, which is subscribed to ceph in Ubuntu. https://bugs.launchpad.net/bugs/1056462 Title: The

[Bug 1056103] Re: no docs package / html docs

2012-09-29 Thread Sage Weil
I think if we package them at all, they should probably be in the regular debian directory so that they get built along with regular releases... otherwise it will be a lot of additional work to build and publish the packages. The rules or Makefile would probably need to do things like clone

[Blueprint servercloud-p-ceph] Preparing CEPH for main inclusion

2011-10-18 Thread Sage Weil
Blueprint changed by Sage Weil: Whiteboard changed:  - MIR librbd and librados  - compile qemu-kvm with ceph (rbd, rados) block device support, see: - * https://launchpad.net/~kirkland/+archive/virt/+packages +   * https://launchpad.net/~kirkland/+archive/virt/+packages + - include latest