Hello,

We have an Avid ISIS 5500 v4.7.5.15920.

Coming up to this last weekend there was a drive in the ISIS that was racking up many errors (over 2000 current pending sector counts in the Storage Manager Log Viewer diskSmart logs) and we were getting very regular 'Long Disk IO on DATA READ' warnings from the drive group the failing drive was located in.

Then finally the ISIS Management Console started displaying a Disk Performance Degraded message so I thought it was time to replace the failing drive on Saturday.

I've replaced the failing drive and now everything has rebuilt fine and we don't seem to be experiencing any issues.

However, the Disk Performance Degraded message persists even though the offending drive has been removed.

I've checked and there's two other drives in the ISIS that are logging very minimal errors (<= 2 uncorrectable sector counts in the Storage Manager Log Viewer diskSmart logs) and the amount of errors on them have not increased for some months. Plus we're only just now slowly starting work again this morning so it's not as though there's a lot of load on the ISIS at the moment.

I'm wondering, how does one clear the Disk Performace Degraded message? Do I need to restart the ISIS or will the message clear after some time under its own volition?

Thank you very much.

Hi,

The message should clear by itself anywhere between immediatly when the rebuild completes and about 3-4 hours of use. I monitor the storage manager disk status of each raid set and monitor the media error count in the megaraid utility every 1-2 hours if I see the disk performance degraded.

I also install hdsentinel to monitor drive health. While I leave it installed I only run it during a check and never leave it running in the background with status in the task bar although I have not seen any side effects except the megaraid storage manager reporting unexpected sense. I do this because hdsentinel gives more direct user friendly information and also helps me predict drive failures.

Thank you for this.

The message still persists yet MegaRaid shows everything as working optimally with no errors and the amount of errors in the storage manager disk status have not increased since before replacing the failing disk. I think I will try hdsentinel. Is there a particular version of it you usually use? It could also be that the error is just taking a little longer than usual to clear and I’m being paranoid, but I’d prefer to be more safe than sorry.

You can reboot it but a warning.... Rebooting a very old ISIS can lead to other problems. It is like gutting an old house and discovering problems behind the wall and under the floors.

 

 

I absolutely understand you, I am trying to resolve the issue without resorting to a restart since we're in the middle of production right now. I may just live with the error since nothing seems to be going wrong at the moment.

Dom while I have you, I know that when we first got the ISIS it was recommended that we restart it once a month just for maintenance purposes. Would you say now that I should probably avoid restarting it at all since the ISIS is quite long in the tooth?

Thank you.

Once a month was a bit excessive in my opinion, especially when the server is busy 18 hours a day / 7 days a week. Normally, I recommend every 2-3 months. Although, we had servers stay on for months before a reboot occured.

On an older ISIS, I would say reboot only when needed. Rebooting is a stressful process for the hardware. At certain point, you dont want to add stress. Just monitor your hard-drives. Take 5 min to check the drives' SMART logs 2 to 3 times a week.

 

Got it. 

Thanks very much for your insight.