9fans - fans of the OS Plan 9 from Bell Labs
 help / color / mirror / Atom feed
* [9fans] strange complete hangs after a fossil snap
@ 2007-01-31 21:07 Robert Raschke
  2007-01-31 22:47 ` Robert Raschke
  2007-01-31 23:03 ` Russ Cox
  0 siblings, 2 replies; 4+ messages in thread
From: Robert Raschke @ 2007-01-31 21:07 UTC (permalink / raw)
  To: 9fans

Hi,

ever since I recently pulled all the newly rebuilt binaries and
managed to rebuild my pccpuf kernel I see odd behaviour whenever my
fossil snaps.  That is both the regular snapshots and the archival
snapshots seem to complete correctly, but immediately afterwards my
cpu machine freezes up.  It feels very much like the file system is
frozen, since some shell builtin commands appear to work, but as soon
as I do something file based the whole thing freezes without any
indication as to what is going on anywhere.

Does anyone see similar behaviour?

Thanks for any pointers in trying to get to the bottom of this,
Robby


--
r dot raschke at tombob dot com




^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [9fans] strange complete hangs after a fossil snap
  2007-01-31 21:07 [9fans] strange complete hangs after a fossil snap Robert Raschke
@ 2007-01-31 22:47 ` Robert Raschke
  2007-01-31 23:03 ` Russ Cox
  1 sibling, 0 replies; 4+ messages in thread
From: Robert Raschke @ 2007-01-31 22:47 UTC (permalink / raw)
  To: 9fans

I wrote:
> ever since I recently pulled all the newly rebuilt binaries and
> managed to rebuild my pccpuf kernel I see odd behaviour whenever my
> fossil snaps.  That is both the regular snapshots and the archival
> snapshots seem to complete correctly, but immediately afterwards my
> cpu machine freezes up.  It feels very much like the file system is
> frozen, since some shell builtin commands appear to work, but as soon
> as I do something file based the whole thing freezes without any
> indication as to what is going on anywhere.
>
> Does anyone see similar behaviour?
>
> Thanks for any pointers in trying to get to the bottom of this,
> Robby

Oh well, it did happen three times in a row, but now everything is
fine.  I am wondering if it was just me being impatient with the
snapshots finishing due to the amount of recent changes that I pulled.
Probably.

Apologies for the noise,

Robby

--
r dot raschke at tombob dot com




^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [9fans] strange complete hangs after a fossil snap
  2007-01-31 21:07 [9fans] strange complete hangs after a fossil snap Robert Raschke
  2007-01-31 22:47 ` Robert Raschke
@ 2007-01-31 23:03 ` Russ Cox
  2007-01-31 23:04   ` Russ Cox
  1 sibling, 1 reply; 4+ messages in thread
From: Russ Cox @ 2007-01-31 23:03 UTC (permalink / raw)
  To: Fans of the OS Plan 9 from Bell Labs

> ever since I recently pulled all the newly rebuilt binaries and
> managed to rebuild my pccpuf kernel I see odd behaviour whenever my
> fossil snaps.  That is both the regular snapshots and the archival
> snapshots seem to complete correctly, but immediately afterwards my
> cpu machine freezes up.  It feels very much like the file system is
> frozen, since some shell builtin commands appear to work, but as soon
> as I do something file based the whole thing freezes without any
> indication as to what is going on anywhere.

This is sad but true.  I have explained it on 9fans before, but I cannot
find the messages.  Fossil's block locking needs to be reworked some
so that blocks waiting to be written (which are locked for that purpose)
can still be locked and copied.  I think a simple read-write lock might
suffice, but I am not doing active fossil development anymore.

Russ


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [9fans] strange complete hangs after a fossil snap
  2007-01-31 23:03 ` Russ Cox
@ 2007-01-31 23:04   ` Russ Cox
  0 siblings, 0 replies; 4+ messages in thread
From: Russ Cox @ 2007-01-31 23:04 UTC (permalink / raw)
  To: Fans of the OS Plan 9 from Bell Labs

> This is sad but true.  I have explained it on 9fans before, but I cannot
> find the messages.  Fossil's block locking needs to be reworked some
> so that blocks waiting to be written (which are locked for that purpose)
> can still be locked and copied.  I think a simple read-write lock might
> suffice, but I am not doing active fossil development anymore.

I should add that the length of wait is in fact proportional to the
amount of data being flushed to venti, and that the bug was not
recently introduced but has always been there.  You saw a big delay
because there was a lot to write to venti, not because fossil got
worse.

Russ


^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2007-01-31 23:04 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-01-31 21:07 [9fans] strange complete hangs after a fossil snap Robert Raschke
2007-01-31 22:47 ` Robert Raschke
2007-01-31 23:03 ` Russ Cox
2007-01-31 23:04   ` Russ Cox

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).