[ZODB-Dev] Transaction Manager Changes

[Toby Dickenson]

On Tuesday 26 November 2002 3:52 pm, Jeremy Hylton wrote:
> >>>>> "TD" =3D=3D Toby Dickenson <tdickenson@geminidataloggers.com> wri=
tes:
>
>   TD> I am seeing some unusual behaviour in some classes which hook
>   TD> into the transaction manager since Jeremy's sortKey changes. I
>   TD> have not been able to reproduce a minimal test case yet, so this
>   TD> has been diagnosed from activity logs on a production system.
>
>   TD> The following two call sequences are ringing my alarm bells:
>
>   TD> 1. abort, abort
>
>   TD> 2. tpc_begin, abort, tpc_abort
>
>   TD> Does this indicate a bug somewhere in the transaction manager,
>   TD> or should my objects be expecting this?
>
>   TD> (if it helps, there may be subtransactions involved too)
>
> I don't have a good answer off the top of my head.  When I wrote the
> first version of deadlock prevention code, I got a number of things
> wrong in the subtransaction support.  I fixed everything that was
> caught by the test suite.  If you see new failures in an application,
> my first guess would be bugs in the code that weren't covered by the
> test suite.
>
> I'm not sure if I understand the problem report.  Are the call
> sequences you report what is happening at the jar, e.g. the jar is
> getting its abort() method call twice in a row?

Yes.

[I have touble calling these objects 'jars'. They are an object that is f=
ound=20
in the _p_jar attribute of an object registered with the transaction mana=
ger.=20
=46rom the Transaction's point of view I guess that makes them a jar]