Based only on code review, not looking at the artifacts, +1. Despite it being 3 years old, I'm a little worried about GERONIMO-6372, the implementation seems to assume we are processing < 1 tx per millisecond. Maybe we should shift the SystemTimeMilliseconds over a few bits just in case hardware gets even faster?
thanks david jencks On Apr 17, 2015, at 7:24 AM, Guillaume Nodet <[email protected]> wrote: > I've uploaded a 3.1.2 release of the transaction manager component at > https://repository.apache.org/content/repositories/orgapachegeronimo-1018 > > This fixes the following issues: > * GERONIMO-6541: NPE during XAResource recovery with invalid Xids > * GERONIMO-6372: RecoveryImpl completing in-progress transactions, > XidFactoryImpl needs to be smarter with matching > * GERONIMO-6373: Expose HOWL flushPartialBuffer config via HOWLLog - useful > under low concurrency > > Please review and vote ! > > Cheers, > Guillaume Nodet >
