On Feb 8, 2013, at 12:21 AM, Adam Goryachev [off-list ref] wrote:

I changed these two settings last night (noop and nr_request = 1) and
today seemed to produce more complaints, and more errors logged about
write failures, so I have restored nr_requests to 128 and the scheduler
back to deadline.

Specifically what errors are being logged? And if this is on clients, what are you getting in dmesg on the storage server at the time of these errors? There might be a lot.


Chris Murphy