On 05/22/2012 01:17 PM, Daniel Schütze wrote:
> Can anyone suggest to me what might be going wrong, or if there is
> anything useful I can examine? So far this has happened 7 times up to
> physmessage_id 612844 (out of the 900k or so) so it is infrequent but
> being a coredump it does mean it halts the migration which is
> inconvenient. I am running MySQL 5.1.55, on freebsd 8.2 for general
> information.
Did you run the full schema migration successfully first?
> Also, and separately, the migration is pretty slow overall which is an
> issue for the downtime it will cause. Looking at the test box (4 core
> Xeon, 4gig ram, standard hdd) during the conversion the processor load
> is low (20% overall i.e. 1 processor at under 100%), the disk utility is
> low (normally under 50%) and innodb pool is set to 500meg which is
> obviously full, so can someone suggest what is rate limiting the conversion?
Running dbmail-util -M is not required and does not require downtime.
The data-store is backward compatible - only body searches will not work
on messageblks.
That said: I haven't seen this problem yet.
Since the core dump seems to originate from gmime, I would advise you to
upgrade to the latest gmime-2.4. I see rfc2047 related bugfixes at least
in 2.4.30, and probably later ones as well.
--
________________________________________________________________
Paul J Stevens pjstevns @ gmail, twitter, skype, linkedin
* Premium Hosting Services and Web Application Consultancy *
www.nfg.nl/[email protected]/+31.85.877.99.97
________________________________________________________________
_______________________________________________
DBmail mailing list
[email protected]
http://mailman.fastxs.nl/cgi-bin/mailman/listinfo/dbmail
