sieve parse error problems

[Gary Windham]

Hi all,

I'm running 2.1.14 on Redhat Linux 7.3 w/ reiserfs for the mailbox and config
partitions.

We recently had some hardware problems with our disk array, had to run a
'reiserfsck --rebuild-tree' against our config partition, and had to perform a
few reconstructs on some users' mailboxes.  Everything appears to be running
fine now, with the exception of a large number of "sieve parse errors" in the
Cyrus log, such as as follows:

Mar 12 12:45:49 smeagol-hb lmtpd[19356]: sieve parse error for devra: line 1:
parse error, unexpected $undefined.
Mar 12 12:46:25 smeagol-hb lmtpd[19356]: sieve parse error for rnr: line 1:
parse error, unexpected NUMBER
Mar 12 12:46:45 smeagol-hb lmtpd[19356]: sieve parse error for smetzger: line 1:
parse error, unexpected $undefined.

I've read the archives and am aware of bug #1951 (Sieve lexer/parser not
resetting its state after a runtime failure).  My problem is, this _never_
happened prior to this hardware issue.  I have gone through the logs and, every
time this problem occurs I find the very first instance of a parse error for
the particular lmtpd process in question.  None of the sieve scripts for these
"first instances" have been corrupt--they all look fine and run fine through
the sieve/test utility.

So, my question is there any way I can determine what is initially causing the
sieve code to get into a bad state?  Nothing in the log indicates where the
initial failure is occuring.

Thanks for any and all assistance,
--Gary
--
Gary Windham
Systems Programmer, Principal
The University of Arizona, CCIT

---
Home Page: http://asg.web.cmu.edu/cyrus
Wiki/FAQ: http://cyruswiki.andrew.cmu.edu
List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html