I was told that this changed with 5773 (the masking while locked behavior) by EMC support a while ago.

As to whats holding your lock - no idea there   Sorry.

Strangley , I have just experienced the same at a customer site.

The lock appeared to be against a GateKeeper -  this was a SAN management server only and had no other connection to the storage device, plus the error in ECC pointed to this server only, possibilly not definitive prove but gave me a starting point.

Although the change only went through yesterday previous attempts to remove the lock worked, but they returned the next day initiated it seemed by a sym discovery action.

When looking in ECC alerts I could see "A Thread in the Storage Agent for Symmetrix on host mickey_mouse has hung. It may become necessary to restart the agent"

The customer has carried this out along with a config change in the options file:

c:\pro*\emc\symapi\config\options

In here the config on the parameter option SYMAPI_WAIT_ON_LOCKED_GK was changed from disabled to enabled.

Unfortunatley the two changes were made at the same time so I cannot say which one "appears" to have resolved the problem, my thoughts are the restarting of the agent was all that was needed, I have mentioned both as both were changed, as mentioned the issue returned the next day for the customer but it appears to have gone for now and it is being monitored to ensure the issue is resolved.

I also thought that Eng code 5773 resolved this but releases after this are where the resolution appears, this customer appears to be pre this release.

Trusts this helps

Thank you for your reply.  This is our target SRDF array and though it does have some hosts attached actively using disk, we rarely need to modify anything.  I will keep the agent and options file items in mind.  I also plan to add more gatekeepers to the host (I also get the error on a managment-only host).