[Bucardo-general] Problem with Bucardo tests

[Greg Sabino Mullane]

On Wed, 27 Feb 2008 16:21:17 +1030
Giulio Harding <giulio.harding at mnetcorporation.com> wrote:

> Hmm, increasing those timeouts didn't help much (despite being an  
> order of magnitude greater) - the tests still fall over at various  
> different points. If anything, there's even less consistency - in  
> some cases, the tests are getting further along (up to the 500s) but  
> in other cases it's still failing early. I'm not yet at the point  
> where I understand what's going on in the tests (beyond some basic  
> aspects), so I'm not sure how to interpret some of the errors I'm  
> seeing, *but* the one thing that does seem to be possibly consistent  
> is a report of some mismatch between some data structures that are  
> being copied & compared, e.g.
...
> #   Failed test '(fullcopy) Second table bucardo_test2 got the  
> fullcopy row [line: 2600]'
> #   at t/01bc.t line 1188.
> #     Structures begin differing at:
> #          $got->[0] = Does not exist
> #     $expected->[0] = ARRAY(0x18c4ff0)
> 
> Are they related?

No, the above just means that the copy was not made to bucardo_test2, and
no rows were returned when we selected the rows that should have been
there.

>  From the output + log.bucardo file that the tests generate, I have  
> attached some snippets below... any other ideas? I'm somewhat  
> reluctant to go ahead with testing Bucardo if I can't get the tests  
> to work :( What's the ETA on the new tests?

Sorry, no ETA yet, except perhaps by the end of the month. I'd like it to
be sooner, but I've no time in the foreseeable next couple weeks to work
on it. I looked through your logs and did not see anything unusual, but
I've found through hard experience that the tough problems never reveal
themselves easily.

All of these tests are a little contrived - you might try a more "real
world" test using Bucardo in the way you envision using it. At the very
least, some failures there might help us find the problem and allow us to
write better tests. :)


-- 
Greg Sabino Mullane greg at endpoint.com
End Point Corporation 610-983-9073
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 163 bytes
Desc: not available
Url : https://mail.endcrypt.com/pipermail/bucardo-general/attachments/20080304/f4ca1235/attachment.bin