./osd/OSDMap.h:460: FAILED assert(exists(osd) && is_up(osd))

2010-10-26 Thread Christian Brunner
When accessing multiple RBD-Volumes from one VM in parallel, we are receiving an assertion: ./osd/OSDMap.h: In function 'entity_inst_t OSDMap::get_inst(int)': ./osd/OSDMap.h:460: FAILED assert(exists(osd) && is_up(osd)) ceph version 0.22.1 (commit:c6f403a6f441184956e00659ce713eaee7014279) 1: (Obje

osd/ReplicatedPG.cc:2403: FAILED assert(!missing.is_missing(soid))

2010-10-26 Thread Christian Brunner
Here is another problem (I think it's unrelated to the previous, but I'm not sure). One of our osds crashed with the following message: osd/ReplicatedPG.cc: In function 'void ReplicatedPG::sub_op_modify(MOSDSubOp*)': osd/ReplicatedPG.cc:2403: FAILED assert(!missing.is_missing(soid)) ceph version

Re: ./osd/OSDMap.h:460: FAILED assert(exists(osd) && is_up(osd))

2010-10-26 Thread Gregory Farnum
Do you have any down OSDs? Looking through the code I am noticing an issue with down OSDs where op_submit can get an error code that it isn't watching for, but there may be checks in other places to prevent that actually happening. -Greg On Tue, Oct 26, 2010 at 5:13 AM, Christian Brunner wrote: >

unstable branch: ls -l hang after write by 64 clients

2010-10-26 Thread Jim Schutt
Hi, I'm testing a file system with 10 osd daemons, 1 mds daemon, 1 mon daemon. Client code is 2.6.36 + current for-linus branch Server code is unstable branch, 2a3e73bb325f. After mounting 64 clients, I had each write its own 512 MiB file with dd, with all files being created in the same direct

Re: ./osd/OSDMap.h:460: FAILED assert(exists(osd) && is_up(osd))

2010-10-26 Thread Sage Weil
On Tue, 26 Oct 2010, Christian Brunner wrote: > When accessing multiple RBD-Volumes from one VM in parallel, we are > receiving an assertion: > > ./osd/OSDMap.h: In function 'entity_inst_t OSDMap::get_inst(int)': > ./osd/OSDMap.h:460: FAILED assert(exists(osd) && is_up(osd)) Can you change that i

AW: ./osd/OSDMap.h:460: FAILED assert(exists(osd) && is_up(osd))

2010-10-26 Thread Christian Brunner
Yes, one OSD crashed approximately an hour before this was happening. Christian -Ursprüngliche Nachricht- Von: gr...@hq.newdream.net [mailto:ceph-devel-ow...@vger.kernel.org] Im Auftrag von Gregory Farnum Gesendet: Dienstag, 26. Oktober 2010 20:00 An: Christian Brunner Cc: ceph-devel@vge

AW: ./osd/OSDMap.h:460: FAILED assert(exists(osd) && is_up(osd))

2010-10-26 Thread Christian Brunner
I can't promise this for tomorrow, but I think I can do this on Thursday. Christian -Ursprüngliche Nachricht- Von: Sage Weil [mailto:s...@newdream.net] Gesendet: Dienstag, 26. Oktober 2010 21:09 An: Christian Brunner Cc: ceph-devel@vger.kernel.org Betreff: Re: ./osd/OSDMap.h:460: FAILED

AW: osd/ReplicatedPG.cc:2403: FAILED assert(!missing.is_missing(soid))

2010-10-26 Thread Christian Brunner
Does someone know which commit this is? I don't want to switch to the unstable branch at the moment. Christian -Ursprüngliche Nachricht- Von: Smets, Jan (Jan) [mailto:jan.sm...@alcatel-lucent.com] Gesendet: Dienstag, 26. Oktober 2010 14:37 An: Christian Brunner Betreff: RE: osd/Replica

Re: unstable branch: ls -l hang after write by 64 clients

2010-10-26 Thread Sage Weil
On Tue, 26 Oct 2010, Jim Schutt wrote: > Hi, > > I'm testing a file system with 10 osd daemons, > 1 mds daemon, 1 mon daemon. > > Client code is 2.6.36 + current for-linus branch > Server code is unstable branch, 2a3e73bb325f. > > After mounting 64 clients, I had each write its own 512 MiB > fi

ceph and large xattrs

2010-10-26 Thread Yehuda Sadeh Weinraub
We recently hit an issue with xattrs on different underlying filesystems. The basic problem is that we use xattrs for various metadata information that we attach to the different objects, and for some cases it tends to grow to relatively large sizes. It appears that the various underlying filesyste

Re: unstable branch: ls -l hang after write by 64 clients

2010-10-26 Thread Jim Schutt
Hi, On Tue, 2010-10-26 at 13:25 -0600, Sage Weil wrote: > On Tue, 26 Oct 2010, Jim Schutt wrote: > > Hi, > > > > I'm testing a file system with 10 osd daemons, > > 1 mds daemon, 1 mon daemon. > > > > Client code is 2.6.36 + current for-linus branch > > Server code is unstable branch, 2a3e73bb325f

unstable branch: you may need to re-run automake

2010-10-26 Thread Colin McCabe
Hi all, I moved a file, and added anew GTK-related macro. So you may need to re-run autogen.sh the next time you pull the unstable branch. If you get an error about not being able to find ceph.o, you have hit this. cheers, Colin -- To unsubscribe from this list: send the line "unsubscribe ceph-d