On Wed, 25 May 2011, Andre Noll wrote: > On Wed, May 25, 09:16, Sage Weil wrote: > > > It's trivial to reproduce: Executing "ls -l" on one client while > > > another client is writing causes the ls command to hang indefinitely. > > > > This was recently fixed in the master branch, commit > > ee7fa813ef29890557f0b03bd3950d422484215d. It hasn't made it into any > > releases yet. I'll cherry-pick it into stable for 0.28.2. > > > > Please let me know if you continue to see this problem with that fix in > > place. > > Nope. I'm now running the ceph master branch (fe955881) on all nodes > but the problem remains. Even after stopping the writer, the "ls -l" > hangs. Linux kernel version is plain 2.6.38.7 everywhere and the ceph > setup is fairly simple: 6 cosds, one cmon and one cmds. I created a new bug (http://tracker.newdream.net/issues/1110) for this. Can you generate an mds log for this and attach it to the bug? debug mds = 20 debug ms = 1 in the [mds] section and restart, or if the cluster is already up just ceph mds tell 0 injectargs '--debug-mds 20 --debug-ms 1' and then reproduce the problem. Should be at /var/log/ceph/mds.a.log or similar. Thanks! sage -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
