[Bucardo-general] bucardo tables causing REINDEX fail

[Greg Sabino Mullane]

On Sun, Apr 08, 2012 at 07:51:43AM -0400, Jonathan Brinkman wrote:
> Greetings
> For the second Sunday AM in a row my weekly REINDEX failed. It now fails
> when I run the REINDEX manually. 
> It is due to 2 bucardo tables locked, apparently. How can I fix this so the
> weekly routine succeeds?
...
> 120249365 | bucardo_delta
> 120253675 | bucardo_delta_archive_log_gateserviceerrors_txn

Hm...are you running a mixture of B4 and B5?

I'm not sure offhand what can be done, as REINDEX needs a heavy lock and 
those tables are obviously very busy. If you can stop things from 
writing to the gateserviceerrors table while the reindex is going on, 
that might help. And switching to B5 is the only thing that will 
prevent bucardo_delta from causing problems, as it is shared across 
all tables in B4, so you would need to stop everything to ensure it 
gets reindexed.

-- 
Greg Sabino Mullane greg at endpoint.com
End Point Corporation
PGP Key: 0x14964AC8
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 163 bytes
Desc: not available
Url : https://mail.endcrypt.com/pipermail/bucardo-general/attachments/20120410/3654b0cc/attachment.bin