[openib-general] RFD: uverbs and hotplug

[Grant Grundler]

On Tue, May 03, 2005 at 08:53:11PM -0700, Caitlin Bestler wrote:
> And applications should be prepared for file systems to fail as well,
> but that doesn't mean that the file system shouldn't consider features
> such as journaling that minimize how often the application will be
> faced that sort of problem.

Sorry - journaling seems like a bad example. Journalling is how the OS
recovers/restores sane state to the filesystem and is not a mechanism
that helps the application recover.

Applications normally depend on another layer to "tolerate" 
storage link failures - aka multipath or mirroring. But it's not
any form of recovery either. There's alot more history
behind the assumption that storage has a reliable transport.
Networking never had that assumption though TCP faked it
for the most part.

> It would certainly be valid to decide that the effort was not worthwhile
> given the effort required. But such a decision should be made with
> full awareness that the notice will enable very few applications to
> actually recover when an entire HCA is taken down.

Agreed.

thanks,
grant

> If the goal
> is to realistically allow applications to confinue functioning after
> the loss of multiple ports some more powerful hooks are required.
> _______________________________________________
> openib-general mailing list
> openib-general at openib.org
> http://openib.org/mailman/listinfo/openib-general
> 
> To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general