Yeah - unfortunately I don't know why either, but people see assorted bizarre
crashes in buffer operations
> userland nfs help the situation. That was basically my next step in
> things to try.
Please do - that eliminates knfs itself from the enquiry potentially.
> reason being support for the Adaptec 2940 U2W cards. At the time, it
> seemed at if the U2W was not supported by the driver in 2.0.36. Was I
> wrong? Plus I fear changes in raid may make things rather difficult to
2.0.36 and 2.2.0 are basically the same aic7xxx driver. 2.0.37 and 2.2.2
will be basically the same newer aic7xxx driver version.
> move back. This is unfortunately a production level machine. Downtime to
> backup, wipe out the partition and rebuild would well, hurt.
Nod
> Next time it dies I will do this and get a copy and send it to the list.
> Is there any way to anticipate a failure, or something to watch that will
> tell me something is wrong?
You might want to try dumping ps -edalf to a log file every 2 minutes or so
that way there is a small chance you'll see a whole pile of jammed up
processes just before it dies
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/