Subject: 'tail -f' and NFS
To: None <firstname.lastname@example.org>
From: Wolfgang Rupprecht <wolfgang+gnus20020912T051735@wsrcc.com>
Date: 09/12/2002 05:30:09
I've got a tail -f running during a 5-day long rebuild of pkgsrc on a
-current sparc ss5.
# tail -f build-pkgsrc-sparc.log
[lots 'o output] ...
tail: build-pkgsrc-sparc.log: Input/output error
# ll build-pkgsrc-sparc.log
32672 -rw-r--r-- 1 root staff 33430717 Sep 12 05:20 build-pkgsrc-sparc.log
The log file is on an NFS-ed directory running on an i386 machine, and
everyone is running -current as of 9/7/2002.
This exact error w. errno has happened a couple of times now, usually
after I leave the 'tail -f' running for over 12hrs. Is this an NFS
bug (or tail bug)? The compile is also using sources mounted over NFS
and I see no errors inside the log file so this isn't some gross NFS
error where everything goes away.
Anyone know what is going on here?
Wolfgang Rupprecht <email@example.com>
Hackers don't spread viruses. Microsoft spreads viruses.