9fans - fans of the OS Plan 9 from Bell Labs
 help / color / mirror / Atom feed
* [9fans] venti icachewriteproc going past isect
@ 2009-01-07 22:34 Oleg Finkelshteyn
  0 siblings, 0 replies; only message in thread
From: Oleg Finkelshteyn @ 2009-01-07 22:34 UTC (permalink / raw)
  To: 9fans

hi,

i've had this behavior some time ago and haven't solved that, and now
Akshat Kumar is experiencing exactly the same, but as the venti at
faul is running on his mail server, he's unable to post here.

as soon as venti starts, an icachewriteproc can be observed which
produces high load, making the whole system unresponsive:

bootes           13    0:00   0:27    42392K Pwrite   venti
[icachewriteproc:/dev/sdC0/isect]

from /proc/13/fd it can be seen that its offset into /dev/sdC0/isect
keeps increasing, even past the partition size:

% ls -l /dev/sdC0/isect
--rw-r----- S 0 bootes bootes  3175279616 Aug 22 23:38 /dev/sdC0/isect
% cat /proc/13/fd
8 rw S    0 (0000000004300066 1 00)     0 3223880192 /dev/sdC0/isect

this always reproduces after a restart.

killing icachewriteproc renders fossil unresponsive, essentially
making system unusable.

fshalt -r hangs at 'syncing.../srv/fscons...', so the machine can't be
rebooted cleanly.

any hints?

thanks,
oleg



^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2009-01-07 22:34 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-01-07 22:34 [9fans] venti icachewriteproc going past isect Oleg Finkelshteyn

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).