Currently, I've got a kernel compiled with the 2003-04-07
patchset (excluding the kdb-i386 patch). It's been running for
two days, and just now it got two processes stuck in 'D' state.
The first process was trying to rm -Rf a directory. The second
process is trying to du -hs that same directory (I got impatient
and suspicious when the rm -Rf sat there for an hour or so and
decided to check on progress). Apparently, whatever problem is
occurring, it's following a specific inode in said directory,
because I can still traipse about the rest of the fs with no
problems.
Until today, this FS had no problems. Until two days ago, it was
running the 1.0.2 release on a 2.4.18 kernel. The system itself
is an EV56/PC164LX. The kernel itself, aside from the xfs
patchset, has the ptrace patch applied, plus a minor patch
(attached) to fix a silly little compilation bug. (Pardoning my
very blunt and rather pointed question, do these patchsets even
get a test compile?)
Looking back over recent linux-xfs@ archives, this seems to be a
recurring problem lately?
--
Kelledin
"If a server crashes in a server farm and no one pings it, does
it still cost four figures to fix?"
kernel-2.4.20-1-xfs_types.patch
Description: Text Data
|