Start a Conversation

Unsolved

This post is more than 5 years old

2524

December 9th, 2015 16:00

Cannot remove Failed device

I have a failed device and I want to remove it.  For what ever reason, it keep telling me "Remove Pending".  I have no idea what to do at this point.  The disk is dead and I can't remove it from the volume.

51 Posts

December 11th, 2015 12:00

Tanr,

Have you attempted to abort the device removal?  Give that a shot and see what the device status is afterward. Reattempt removal if it will let you.

You could also try marking the disk devices as active in the OS, and see if it will let you remove the device then.

Let me know how that goes.

34 Posts

December 15th, 2015 10:00

What version of ScaleIO are you using?

Is this on ESXi, Linux, Windows or Bare Metal?

Did you use the "remove_sds_device" scli command and it now has you pending?

Have you gone out to the support.emc.com web site for Disk Replace?

Please supply us with a bit more information about your setup so we can help you better.

Thank you

9 Posts

December 16th, 2015 16:00

I tried both from GUI and scli command (Linux).  When I use the GUI, it keep showing "Remove Pending".  From scli:

# scli --clear_sds_device_error --sds_ip --clear_all

Successfully cleared all SDS with IP devices

But then I run scli --query_sds, it still showing "1 devices with error"

When I run scli --remove_sds_device, it shows "SDS Device with ID 366df7e000000000 is being removed asynchronously" but it doesn't do anything.  The GUI again shows "Remove Pending".

December 21st, 2015 00:00

1. Try to run: scli --abort_remove_sds_device 1st

2. Clear device error

3. Remove sds device

Keep in mind that you need to have enough Spare to evacuate / rebalance the info from that device to other devices (if indeed you have mapped volumes). Status of remove pending is during the time of evacuating the data from that disk to other disks in the SP, that is why you need sufficient Free space and Spare capacities.

9 Posts

December 22nd, 2015 08:00

@tomer_(engineering_qa), I have already done that.  I gave up on this...  Delete the volume and recreate new one.  At this point, I'm very doubtful if we should use ScaleIO in production. 

No Events found!

Top