[7064] in testers
Re: AFS cache coherency failure under 9.4
daemon@ATHENA.MIT.EDU (Greg Hudson)
Wed Jun 1 12:50:53 2005
Date: Wed, 1 Jun 2005 12:50:28 -0400
Message-Id: <200506011650.j51GoSAj018374@egyptian-gods.mit.edu>
From: Greg Hudson <ghudson@MIT.EDU>
To: Greg Hudson <ghudson@MIT.EDU>
CC: testers@MIT.EDU
In-reply-to: <200506010421.j514LQrx012413@egyptian-gods.mit.edu>
I observed another case of cache coherency failure, this time on
erato:
erato.mit.edu# more /afs/dev/user/ghudson/zephyr/zclients.pl
[old version of script]
erato.mit.edu# fs flushv /afs/dev/user/ghudson/zephyr/zclients.pl
erato.mit.edu# more /afs/dev/user/ghudson/zephyr/zclients.pl
[newer version of script]
Since erato is running a presumably good AFS client, the lossage may
be on the client of my laptop which has been saving the file (which is
also the machine used to save the file in the previous case). Just as
on equal-rites, my laptop's dmesg output has a lot of stuff like:
post_create: no inode, dir (dev=afs, ino=52428827)
post_create: no inode, dir (dev=afs, ino=52428827)
post_create: no inode, dir (dev=afs, ino=52428827)
post_create: no inode, dir (dev=afs, ino=52428827)
post_create: no inode, dir (dev=afs, ino=52428827)
My laptop is runing 9.4.5, which uses a 2.6.9-5.EL kernel, so we still
can't blame the errata Red Hat kernel.
