Hot Spare IO Response Time

Hi

We have not seen this behavior before, can you post a screenshot of the performance chart and the list of disks showing the disk as spare.

Brian

Hi,

Thanks for the picts.  I am going to do a little more research, but a few questions:

1) Is there a failover that occurred recently, where these drives have been pulled into service? 

2) Have you put in new drives for failed ones, and are migrating data back from the spare to those drives?

3) How long has this been the case?  If you switch the timeframe on the chart to 7 days, 30 days, etc., when did this behavior first occur?

Brian

Brian, I believe we've had two disk failures in the last year. I couldn't tell you what hot spare was used, but the migration back to the new disks is definitely over. The last failure was at least two months ago.

If I change the timeframe, the same disks are at the top. We have 4 hot spares and its only the same two hot spares that are showing up with high disk IO. Perhaps these are the disks that were used during the failovers.

Should I open up a case for this? Anyone else experience this?

Yes, go ahead and open a case for this.  Brian

I've opened cases with SolarWinds & EMC. So far, no one has been able to explain the extremely high IO Response Times on these two Hot Spares.

There really should be a way to disable monitoring, within Profiler, for Hot Spares. Does anyone know of a way to do this. I don't see any reason to collect statistics on hot spres.

By comparing the data coming directly from navisphere and the data coming from the SMI-S provider we were able to verify that the data that we are displaying is correct compared to the data from navisphere. 

I do believe the statistics are being displayed correctly within Profiler. We actually had a disk fail last week. The hot spare that transitioned and then the failed disk was replaced. The hot spare that was used is no longer showing extremely high disk IO.

I could mention this in the feature request, but it would be nice to have the option to exclude certain disks within graphs, specifically Hot Spares.

Thanks for all your help.

Thanks for the feature request, we will put in in the queue.  Brian