On 11/29/2010 10:52 AM, Chris Conn wrote:
As for your earlier errors... There is a bug with sync_crc calculation in 2.4.4, so I would recommend waiting for 2.4.5 before upgrading again. I'm hoping we can release 2.4.5 very soon, and Kolab will be building their RPMs pretty quickly too I imagine.Hello, I'm hoping that this version will fix my issue with replication I posted to the list a few days back. I am going to build 2.4.5 today. What I am seeing is that replication goes bonkers at 04:02 when the servers replicating to each other send their logrotate messages, I start having CRC check issues, no mailbox found errors and everything barfs. Both servers are sending that inane log scan message to the same mailbox, whatever the root alias is. What is most interesting is that the replication seems stuck on a message it sync'ed hours ago, trying to append to the existing file that from what I can tell is identical on both servers. Reconstructing does nothing, I have to unlink manually the file in the mailbox in question, reconstruct, and then things start working again.
<snip>
so my only solution was to delete the 16. file under user.user, which was an email that was successfully replicated at 00:15 am that same day. So is this the same bug? In any case, this happens every day when the logrotate emails get sent from master to slave and slave to master, as they are sending a different email to the same mailbox roughly at the same time. Hopefully running 2.4.5 will avoid this?
I didn't think that the slave was supposed to be setup to accept incoming mail?
http://comments.gmane.org/gmane.mail.imap.cyrus/33805
Chris ---- Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
<<attachment: boutilpj.vcf>>
---- Cyrus Home Page: http://www.cyrusimap.org/ List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/
