(Illustration by Gaich Muramatsu)
FYI: I just upgraded stat from 2.5 to 3.3 and now we get sane values. # stat . File: "." Size: 2048 Blocks: 4 IO Block: 8192 Directory ..Patrick On Thu, 2005-07-28 at 12:25 -0600, Patrick Walsh wrote: > > > # stat . > > > File: "." > > > Size: 2048 Blocks: 4 IO Block: > > > -4611713506218074112 Directory > > > Device: ah/10d Inode: 1352058954 Links: 2 > > > > This linkcount states that the directory has no subdirectories. So find > > will consider everything a file. > > How can you tell that? "Links: 2"? > > > You could do "cfs er .", which should tell venus to send an invalidation > > call to the kernel. If you then run stat again it should show a > > linkcount of 1 and the find should work again. > > Yes, that is exactly what happened. But the IO Block remained the > same. Actually, it didn't remain the same, but it's still not what you > would expect. Here's an excerpt: > > # stat . > File: "." > Size: 2048 Blocks: 4 IO Block: > -4611716804752957440 Directory > Device: ah/10d Inode: 1352058954 Links: 2 > ... > # cfs er . > # stat . > File: "." > Size: 2048 Blocks: 4 IO Block: > -4611718454020399104 Directory > Device: ah/10d Inode: 1352058954 Links: 1 > > > Here are the values from venus.conf that aren't set to defaults: > > realm=abc > cacheblocks=3145728 > mapprivate=1 > masquerade_port=2435 > > And in case it matters, here's server.conf's non-default values from > one of the servers: > > rvm_log="/vice/RVMLOG" > rvm_data="/vice/RVMDATA" > rvm_data_length="524288000" > ipaddress=x.x.x.x > rvmtruncate=5 > trace=100 > allow_sha=1 > > Do you think the strange block size could indicate a problem or cause a > problem at some point? > > Here's some other info, in case it helps: > > # df /coda > Filesystem 1k-blocks Used Available Use% Mounted on > coda 3145728 464 2937779 1% /coda > # df /vicepa > Filesystem 1k-blocks Used Available Use% Mounted on > /dev/sda3 40313996 1318908 36947204 4% /vicepa > > > On second thought, this might not have anything at all to do with coda. > So don't bother chasing it down. It's either a problem with stat or > with our kernel, is my guess: > > # stat / > File: "/" > Size: 4096 Blocks: 8 IO Block: > -4611715155485519872 Directory > > We're using the latest Redhat AS3 kernel. We're also using a RAID > array, in case that has some impact. Anyway, I think the only problem > here worth tracking down is why on a cp -r (and maybe other operations) > the link count isn't being set properly... > > Thanks again for your help, Jan. > > ..Patrick > > > > -- Patrick Walsh eSoft Incorporated 303.444.1600 x3350 http://www.esoft.com/Received on 2005-07-28 14:37:43