[jboss-dev-forums] [JBoss Transactions Development] - Re: remote tx inflow: timeouts

[Tom Jenkinson]

Tom Jenkinson [http://community.jboss.org/people/tomjenkinson] created the discussion

"Re: remote tx inflow: timeouts"

To view the discussion, visit: http://community.jboss.org/message/630256#630256

--------------------------------------------------------------
As I mentioned above, the fudgefactor is really a transport specific detail. In the normal case the transaction will timeout at the root transaction manager and cascade down to the subordinate transactions managers. Ideally we would not monitor timeout in the subordinate at all and beable to rely on the root TransactionReaper to time us out but if the root (or an intermediary fails) we can't rely on solely on this, hence fudgefactor is there in case a parent transaction manager fails, in which case if the transport can't tell us that a parent has died then we need fudge factor to allow us to say basically we haven't heard anything from the parent in a reasonable time, timeout the transaction here and accept the fact that there will be a heuristic to the parent

I will check what happens in JTS though I am guessing that in JTS the ORB tells you a parent has died so you can clean up your transaction immediately.

Totally appreciate your point about how to determine timeout though, it will need to be a configuration of the transport I would argue.

Ideallly we would be notified by the transport, I am waiting to hear from David whether this is something we can rely on.
--------------------------------------------------------------

Reply to this message by going to Community
[http://community.jboss.org/message/630256#630256]

Start a new discussion in JBoss Transactions Development at Community
[http://community.jboss.org/choose-container!input.jspa?contentType=1&containerType=14&container=2041]

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/jboss-dev-forums/attachments/20111005/b88aad44/attachment.html