Hi,
> I think we should process the unprocessed requests before we have the
> server exit. 
By unprocessed requests you mean unprocessed unexpected messages, job, bmi 
and/or trove requests ? 
I think unprocessed unexp messages shoudn't be processed, the client will 
restart them once the server is back up. Maybe it is useful to finish all 
queued trove requests to ensure metadata consistency ? Due to the internal 
ordering of the trove operations within a sm this should be guaranteed 
somehow even by discarding unprocessed elements ? 
> Server shutdown is not the only reason we have to deal with 
> here. If the server exits with a segfault or bus error or any other
> unclean error, we don't want to be left with a corrupted volume!
Is there a way to intercept the segfault or bus error ?
I thought we couldn't do anything at this point.

Thanks for helping me out,
Julian
_______________________________________________
Pvfs2-developers mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-developers

Reply via email to