Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I am testing out a iSCSI SAN solution in my vSphere environment and have been adding/removing LUNs on the SAN host while some were left connected on ESXi hosts. I had enabled Storage I/O Control on the datastores and now I cannot remove them. The errors received is:

Call "HostDatastoreSystem.RemoveDatastore" for object "datastoreSystem-580" on vCenter Server "VCENTER1" failed.
Cannot remove datastore 'ramdisk' because Storage I/O Control is enabled on it. Correct it and re-try the operation

I cannot disable Storage I/O Control since I cannot right click on the datastore and get to the Properties option. There is a KB article for removing inaccessible NFS datastores:

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2008507

But this doesn't seem to apply to iSCSI datastores. I no longer have the LUN available to represent.

I have also tried unlclaiming the storage device and get an error that it is still busy:

~ # esxcli corestorage claiming unclaim -t device -d naa.600144f0a4a4450000004d7ae2dd0002
Errors:
Unable to perform unclaim.  Error message was : Unable to unclaim all requested paths. Some paths were busy or were the last path to an in use device.  See VMkernel logs for more information.

How can I remove the iSCSI datastore that is missing the LUN?

share|improve this question

3 Answers 3

up vote 1 down vote accepted

Unfortunately, I was forced to reboot this host. I will try contacting VMware about the issue. Though I would guess it is likely a bug on the ESXi 4.1 build I was running.

After a reboot the LUNs were gone and vSphere allowed me to vMotion freely.

share|improve this answer
    
When I attended an advanced SLES11 SP1 course I played a little with iSCSI - same problem there: if the server dies the client won`t get rid of it. This is propably rather a general linux problem than an vmware specific one. Reminds me of early NFS implementations... –  Nils Dec 31 '11 at 22:16
    
Yea, it feels a lot like that. I really like iSCSI. I have started following the linux-iscsi.org project which just recently got accepted into the kernel as of 3.1. Hopefully this will mean some good advances in iSCSI in Linux and in general for 2012. –  Andy Shinn Jan 1 '12 at 6:06

I would try to check if other excli commands apply here http://pubs.vmware.com/vsphere-50/index.jsp?topic=/com.vmware.vcli.ref.doc_50/esxcli_storage.html
For example "storage core device detached list" seems like a good start. This guide may be helpful too:

http://pubs.vmware.com/vsphere-50/index.jsp?topic=/com.vmware.vcli.examples.doc_50/cli_manage_files.5.6.html

share|improve this answer
    
I have actually tried poking around the esxcli and esxcfg-* commands and still get errors removing it. ~ # esxcli corestorage claiming unclaim -t device -d naa.600144f0a4a4450000004d7ae2dd0002 Errors: Unable to perform unclaim. Error message was : Unable to unclaim all requested paths. Some paths were busy or were the last path to an in use device. See VMkernel logs for more information. –  Andy Shinn Dec 26 '11 at 1:33
    
Is there anything useful in the VMkernel logs? –  Sergei Dec 26 '11 at 9:16
    
There is no VMkernel log that I could find in ESXi. But the other logs just show something similar either about Storage I/O Control or the device being in use. I found a command (I forget the command now) that showed me the world ID using the device and it was 4100 which was the idle0 process. This didn't make much sense. But seems to indicate that it was in use by the VMkernel some how. I am being forced to schedule a reboot of this system and will let you know if this fixed it. Unfortunately, this is the least desirable resolution :\ –  Andy Shinn Dec 27 '11 at 19:15
    
It may be worth asking for help at communities.vmware.com too. –  Sergei Dec 27 '11 at 20:23
    
Does this command stop any functionality of the host or vms? –  user188918 Sep 6 '13 at 18:07

I found the KB on the vmware site.

Connect to each ESXi 5.0 host to which the LUN is presented by using SSH. Run this command to stop the SIOC service:

/etc/init.d/storageRM stop

In the vSphere Client, select the host and then click the Configuration tab. Click Rescan All. After the rescan completes, run this command to restart the SIOC service:

/etc/init.d/storageRM start

Note: If the issue persists, put the affected ESXi 5.0 host in the maintenance mode and then reboot the host.

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2011220

share|improve this answer
1  
You should at least quote a section of that link that you think answers the question. –  chutz Oct 25 '12 at 14:51

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.