On Thu, 2008-01-31 at 08:21 -0800, Mike DeMarco wrote: > Have you tried to do a > zpool scrub ? > >
Yes - no joy, completed without errors but didn't fix the problem. The zdb command slowed the entire server down too much, so I'm going to run it over the weekend - this is a production NFS server and the users were getting restless :) > > > Hi all, > > > > I've recently cleared out a directory on a ZFS > > filesystem that had ~2 > > million files in it. Even after clearing out, listing > > the directory > > takes in the order of two minutes. I thought it might > > be something to do > > with the fact that the files were still being held in > > snapshots of the > > the filesystem, so I destroyed these. Listing still > > takes the same > > amount of time; a truss reveals the hanging system > > call to be fstat64(). > > > > .. > > openat(AT_FDCWD, "Trash/", > > O_RDONLY|O_NDELAY|O_LARGEFILE) = 4 > > fcntl(4, F_SETFD, 0x00000001) = 0 > > fstat64(4, 0x08047B10) = 0 > > (hangs for 100s +) > > getdents64(4, 0xFEE14000, 8192) = 144 > > getdents64(4, 0xFEE14000, 8192) = 0 > > close(4) = 0 > > .. > > > > The box is running SXDE 9/07 with ZFS pool version 8. > > I've had a similar > > problem on other versions with filesystems that get > > large numbers of > > files written to and deleted from, e.g. queue > > directories on mail > > servers. > > > > Any ideas why this might be happening, or how to fix > > it? > > > > Cheers, > > > > Chris > > > > > > _______________________________________________ > > opensolaris-discuss mailing list > > [email protected] > > > This message posted from opensolaris.org > _______________________________________________ > opensolaris-discuss mailing list > [email protected] Chris Linton-Ford Systems Administrator Firebox.com t: 0870 420 4475 f: 0870 220 2178 e: [EMAIL PROTECTED] Company number: 3874477 Company registered address: 41 Welbeck Street London W1M 8HD _______________________________________________ opensolaris-discuss mailing list [email protected]
