[Mulgara-general] MulgaraTransactionException

Life is hard, and then you die ronald at innovation.ch
Fri Jul 18 10:44:55 UTC 2008


On Fri, Jul 18, 2008 at 06:29:58PM +0800, Edwin Shin wrote:
> On 07/18/2008 12:58 AM, Edwin Shin is rumored to have said:
> > On 07/18/2008 12:23 AM, Paul Gearon is rumored to have said:
> >> On Wed, Jul 16, 2008 at 10:46 AM, Edwin Shin 
> >> <eddie at fedora-commons.org> wrote:
> >> <snip/>
> >>> Switching to the new branch seems yields different errors, but at least
> >>> allows the unit test to complete. I'm attaching the log output (DEBUG 
> >>> for
> >>> just the classes you listed above).
> >>
> >> Does your latest log include the "errors" you were talking about?
> >>
> >> I ask, because they're not actually errors you can see in the log.
> >> They're stack traces, obtained by instantiating a Throwable and
> >> printing it.
> > 
> > There were errors in stdout. I didn't realize they weren't in the logs I 
> > sent (I should stop sending out emails in the middle of the night). I'll 
> > see if I can reproduce them--the working copy I was using for doing this 
> > testing got a bit hosed earlier today so I may not get to this till 
> > morning.
> 
> I suppose I should be happy to report that I can't seem to reproduce any 
> error messages using the build from the mgr-121-lockrecovery branch. I 
> could have sworn that there were ERRORs logged to the console when I was 
> testing the other night. I might have had some inadvertent mix & 
> matching of java5 and java6 in build & test environments the other day, 
> but that's all I can think of that's different.

See my previous email I just sent. I think you may have some timing
or synchronization issue in trippi. While the mgr-121 branch should
handle this better, I'd still not recommend relying on it.


  Cheers,

  Ronald




More information about the Mulgara-general mailing list