NULL symmaskdb Node and Port Names

Question

I'm trying to determine if there are masked devices that have had the host retired but the devices were never unmasked and reclaimed by the SAN team. I ran a symmaskdb -sid list database and noticed the following entry:

Identifier Type Node Name Port Name Devices
---------------- ----- --------------------------------- ---------
210000e08b9a0c4c Fibre NULL NULL 004F
210000e08b93b09d Fibre NULL NULL 004F

I checked the device in ECC (5.2 SP3) and it showed no relationship to a host so I figured it was unused storage and unmasked the device. Turned out it was one of our ESX server's OS volume

(As an interesting side note the VMs never crashed but the ESX server needed rebooted in ordered to be managed again). After remasking the device the Node name and Port Name showed up as the WWN in the first column. We don't use aliases so there won't be an alias there.

We're in the process of upgrading to ECC 6.0 which does have ESX host agents but until then I'm worried about seeing NULL returned from a symmaskdb command. Why would the Node and Port Name show up as NULL? On a separate storage frame I ran this command again and a different device showed up as NULL with the same WWNs.

210000e08b9a0c4c Fibre NULL NULL 0ECB
210000e08b93b09d Fibre NULL NULL 0ECB

What's weird is that this device is masked to other servers and they don't have this issue. Another oddity is that other devices are masked to these WWNs and they show up fine like:

210000e08b9a0c4c Fibre 210000e08b9a0c4c 210000e08b9a0c4c 0682
210000e08b93b09d Fibre 210000e08b93b09d 210000e08b93b09d 0682

I can't unmask this device to fix the problem but I'm wondering why this occurs and what does or should the NULL entries mean? This is the only other device in our SAN environment that shows up as NULL. How do other people determine if storage can be reclaimed when a host is retired without the SAN admins being notified and without a host agent running?

dynamox · Answer

i do not know why sometimes it shows NULL and sometimes it shows WWN, but i can tell you that you are not alone. If you an open system running Solutions enabler you can run symmask list hba -rename -sid XXX ..and that will replace either NULL or WWN with a host name. For system like ESX where you do not have SE installed, you can manually rename NULL or WWN to a meaningful host name using symmaskdb command. Usually if somebody decommissioned a server and did not let you know , you will see an entree in FA login history (symmaskdb -sid XX list logins -dir 8a -p 0). There you can see either NULL or WWN and next to it in the column 'Login' ..it will say NO. That tells me that the host is either offline, or has been decommissioned ..that's why i like to have meaningful names in symmaskdb so i can tell what that system is or was. So after i find this 'ghost' WWN i run it through symmaskdb -sid XXX -wwn XXXXXX list -devs ...and see what hypers are masked to it, then i check each hyper and see if it's presented to any other WWN other than the one in question ..only then i unmask.

xe2sdc · Answer

Dynamox that's exactly what we do when we make what we call 'spring cleanup' .. Even if it's autumn Message was edited by: Stefano Del Corno

dynamox · Answer

my 'favorite' activity ...man i hate unmasking ..especially on multi-terabytes systems.

Symmetrix

NULL symmaskdb Node and Port Names

Was this post helpful?