On 12/3/19 4:23 pm, Adam Goryachev wrote:
Other things to consider include what to do with the writes intended for this disk while the disk is busy... potentially using the out of sync bitmap is useful here....
And reads. You'll need code that says "this disk is out of service while it is off in the weeds, we need to read every other disk and assemble the data from parity" to avoid the delays we are trying to mitigate.
So now you are queuing writes *and* doing full reconstruction reads for *every* other read that is destined for that disk.
