turbid vale
This caught me by surprise - I expected async delete to have LESS impact on the clients than an xcp delete. I was obviously wrong :(. I ran xcp delete up to 11:30am on this file system, switched to async delete at 11:30, killed it at 17:30, and switched back to xcp at 17:32 where it ran until the next day (total files deleted: >200M!). Is this expected or is this an async delete bug/limitation? Graph from nabox/harvest.
forest saffron
Glad the new dashboard was helpful at least.
I'd open a case because that doesn't sound right Ed.
gusty fern
is this a FAS (spinning disks)?
turbid vale
This was AFF A800. My NetApp SE is all over it. "not designed for performance" is the initial response which of course is not acceptable.
If end-user performance is crippled running async delete, my recommendation is to never use async delete and continue using xcp.
random cobalt
did you log a case for this? I'm curious as to what we found
turbid vale
I did not log a case and have offered the perf archive logs. They wanted a perfstat but I'm not going to reproduce the problem in house (and this may only be caused by the size of the delete - I have no way of knowing). Dan Chuey currently owns it.