On Fri, Sep 06, 2013 at 08:21:57PM -0700, David E. Wheeler wrote:

> get these warnings, which understandably alarm our DBAs:

Alarm them from reading the Bucardo logs? Or in the Postgres logs? 
We can't do much about the latter.

> if (!$g->{has_exception_code}) {
>    $self->glog("Warning! Aborting due to exception for $S.$T:$pkval Error was 
> $err", LOG_WARN);
>    die "$err\n";
> }

> I don’t at all understand what has_exception_code is for

It means if you have no custom code designed to handle any exceptions 
that crop up during the replication.

> but I’m wondering whether, on serialization or deadlock errors, this 
> warning should be suppressed? After all, if I understand the code correctly, 
> that call to `die` is caught by the exception handler which *does* 
> notice serialization and deadlock errors and either retries them or emits 
> an error. Do I have that right? If so, I am not clear that the warning 
> is needed here at all. Is it?

Well, it's only providing some detail to the Bucardo logs, but I could see 
a case for bumping the LOG_x level to something else if it is one of 
the known cases we automatically handle.

-- 
Greg Sabino Mullane [email protected]
End Point Corporation
PGP Key: 0x14964AC8

Attachment: signature.asc
Description: Digital signature

_______________________________________________
Bucardo-general mailing list
[email protected]
https://mail.endcrypt.com/mailman/listinfo/bucardo-general

Reply via email to