Please don't reply to lustre-devel. Instead, comment in Bugzilla by using the
following link:
https://bugzilla.lustre.org/show_bug.cgi?id=11277
What |Removed |Added
----------------------------------------------------------------------------
Attachment #9278|review?([EMAIL PROTECTED])|review+,
Flag| |review?([EMAIL PROTECTED]
| |m)
(From update of attachment 9278)
This patch should ease the pain of this specific problem at LLNL for now.
But as we recently learned, a different measure for entire class of similar
problems was developed for CMD2 project and what we should do is to adopt that
solution.
(do not allow a client to reconnect if there are some requests from it are
processed still).
OBD_RACE change will affect some of our existing regression tests, but should
not affect anything during normal operations.
Andreas, why it was decided not to wake up the other thread in OBD_RACE and
instead only wake it up when second thread finishes its job completely?
_______________________________________________
Lustre-devel mailing list
[email protected]
https://mail.clusterfs.com/mailman/listinfo/lustre-devel
