[sheepdog] Bug in v0.4.0 ? get_sha1_file(80) PANIC: Operation not supported

[Liu Yuan]

On 07/16/2012 05:29 PM, Christoph Hellwig wrote:
> Currently farm always writes a sha1 snapshot file after each recovery
> process.  I can't see how that's needed on a pure gateway, but maybe
> someone who knows farm better can comment?

I have sent a patch for it, that higher level code shouldn't try to do
recovery when it is a gateway only node.

Thanks,
Yuan