We are currently evaluating Jpivot in our web application and the concurrency tests that we made doesnt give us the expected results.
Looking at the source code, we realize that there is a "RequestSynchronizer" that forces jpivot to run one query at once, and that's a death end for us.
We have seen that exits some kind of synch at the httpsession level , and that's ok , but we cant realize the need of don`t support concurrent queries at all.
we could work in the code to modify that behavior but first we want to know if we are missing something.
Thanks in advance,
the RequestSynchronizer is supposed to synchronize requests at the session level, there can be only one request per session at once. This is to prevent that one thread renders a result that is modified by another thread concurrently. There should be no synchronization between sessions, i.e. it should be possible to run two concurrent queries in two different sessions. Everything else would be a bug.
I do think that, the JPivot should restrict the concurrent event to the same model.
but, I wonder, whether the same restriction also appears in the other BI front end?
If still follow the mechanism of the JPivot, it should have the same problem while concurrent(short time continue) calling to the same model.
how can we make it better?
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.