Frank Filz

We are considering options for some level of state persistence to allow transparent state migration.

Is this still an issue of interest? If so, we really need more analysis of what might be going wrong. Our team is pretty small and we have limited capacity...

No response in over 30 days. Closing.

This is unlikely to ever be worked on.

We do have a fix for LRU growth that is in review but that won't land for a few weeks while we try and wrap up V4.3. The fix is...

It would also help to have output from massif that shows the allocation paths that are using the memory in case some other source is at fault.

Could you try with LOG { COMPONENTS { EXPORT = FullDebug; } } That will help diagnose what is causing problems with the PUTFH You could also try V4.2.3 though...

Did you change the export id of your export? It looks like PUTFH is trying to use a no longer valid file handle.

Closing old issues that are not planned for any resolution. Please re-open with new information if this issue is still of concern.

V4.3 has been tagged with a fix for this issue. Closing.