EMC DISK Failures

[Nikee Reddy]

Hello,

Is there any way that we can find out the DISK errors/failures in the EMC symmetrix? Are there any scripts available to monitor the EMC disks status proactively?

Thanks,
Nikee

[Tim Adamson_1]

Hi,

EMC Symmetrix are meant to have the ET capability, ie, phone home!

Contact EMC for further information.

From the OS perspective, you may see errors in syslog and ioscan -fn may show the disks as NO_HW. You may also experience LVM issues too.

Hope it helps

Tim

[Michael Tully]

That's what you pay EMC support for. If there is a problem in the symm, EMC should be calling you and telling you there is a problem. Your waisting your time with scripts, see the logs in /var/adm/syslog and EMS to see problems. 'ioscan' will tell you if you have a connectivity failure by reporting a 'NO_HW' status for the hardware.

[Nikee Reddy]

Hello Guys,

Thanks for the information.

Until now we do have the EMC support and now our management decided not to go with the EMC maintenance due to cost issues and they would like to go-ahead with time and material.

Due to this reason I am looking for some work around methods to figure out the disk failures.


Thanks,
Nike

[Tim Adamson_1]

Hi,

As you don't have the EMC support, you are going to have to use other methods.

Check the syslog file for LVM error messages related to VGs, LVs or PVs on the EMC. This would be your best guide to potential issues with the luns on the EMC.

Also ioscan -fn will provide details if the disks are not available.

Another thing to try would be a dd of the lun and see if you get IO errors (although I believe the EMC Symmetrix will accommodate this sort of problem)

Nothing else springs to mind at the moment.


Tim.

[Con O'Kelly]

Hi
As Tim & Michael said EMC normally has a dial home facility so that you are notified of disk failures.

EMC normally replaces disks long before they actually fail. As I understand it as soon as disk shows a certain number of errors, EMC will notify you & replace the disk before it actually fails.

If an EMC disk does fail, it will show up in syslog.log as any normal disk failure does, but this is after the event.

Your Symmetrix will be using some level of RAID so that you are protected from a disk failure. However you need to constantly monitor syslog.log so that you pick up the disk failure immediately & get it replaced.

Cheers
Con

[Stefan Farrelly]

Install Symmetrix manager on an HP-UX server. You should have got it free when you purchased your EMC. This will allow you to monitor errors yourself in realtime if you run the program, but more importantly set the error level to email root at a particular host if a major problem (like a disk failed or overtemperature). This is what weve done, often we notice first via email from the Symmetrix that a disk has failed before EMC let us know or come in to replace it.