[OpenAFS-port-darwin] Another vnreclaim panic?

[Lidie Steve]

On Aug 10, 2004, at 8:45 AM, Michael G. Ross wrote:

> Was your machine under a Matlab load when the panic occurred? My group 
> seems to be experiencing kernel panics that are

No.  I did the Matlab install/testing noon-ish, yesterday.  Left work 
at 1730-ish.  Panic at 0100 this morning.

> triggered by using Matlab on AFS (although we have the Matlab software 
> installed locally, our .m files and data are all on AFS).
>
> We've experienced this with both Matlab 6 and Matlab 7 on our G5 
> Xserves, unfortunately, our kernel panic logs haven't been informative 
> enough to definitively finger or exonerate AFS (see old messages in 
> the port-darwin list archives).
>
> Recent experience indicates that the problem doesn't happen on Matlab 
> loads that do not touch AFS and just read and write to the local disk, 
> but we aren't totally confident of that analysis yet.
>
> Oddly enough, loads produced by Java and C processes on AFS seem to be 
> completely stable.