does ECC action affect the vcmdb?

ok will check out ther..

yes, it will affect vcmdb ...ECC is your GUI gateway to manage your Symmetrix/DMX ..any masking activity done in ECC will be reflected in vmcdb. Now whether deleting a host object should automatically unmask devices from it ..that probably needs to be asked in ECC forum

I suspect this question may receive better answers in the ECC area of our forums .. This thread will move itself in the next hours

And that's why this thread landed in Resource Management/EMC Control Center area

sorry....

i hope ECC action ( deletion of host container) doent affect the VCM

i mean to refer to this action in ECC.


Looking for the inputs.

Just deleting the host object should not affect the vcmdb.

Can you clarify exactly how you went about deleting the "host container"?

As Allen stated, deleting the host object in the tree panel in ControlCenter Console should have absolutely no effect on the VCMDB. All it simply does is remove the object from the repository database.


Regards,
Séamus Coffey
EMC Global Services

Indeed. As long as you didn't do any unmaskings or anything related to the wwn's and symdevs, you should be fine. Simply add the hosts back to where they were and you're good to go. Since you didn't unmask, adding a host automatically makes the maskings visible again when you click on the host and select masking.

To add some to this if you simply delete the host objects the WWN of the adapters that still have their masking in place should move down to "Unidentified Ports" in the Connectivity folder where they could be accessed to do LUN masking and unmasking.

I saw a rather interesting result to an action I took yesterday that might explain what happened. BTW - This was on CC v6.1...

I was looking at the active zoneset for one of the fabrics in ECC when I noticed that one of the manually created "host" objects had been assigned the wrong HBA ports and IP address. I deleted the object so I could recreate it properly. When I deleted the "host" object the WWN for the associated HBA vanished from the zone in the zoneset I was looking at. I'm not sure if it was a console issue (displayed incorrectly) or if it actually removed the WWN from the zone in that copy of the zoneset in ECC. Unfortunately I was forced to restart the entire environment soon after that and everything looked normal after that.

What I'm wondering, was if I had activated that zoneset at that point in time would it have written out the zone without the WWN in it? That would effectively remove the paths from the host perspective and it would lose it's storage.

I haven't had a chance to try to duplicate this again yet, but it could explain the situation under very specific circumstances.

Interesting, eh?

I would say this was a JAVA glitch. I hope anyway. Is there somebody from the ECC dept here who has an answer on this ? I'd be interested to know what happened here.

Hi Maggie,

I think it'll be impossible to tell without having the logs from the period at hand. The EGN logs would cover any zoning changes and the EGF logs would cover any masking changes. These along with the Server/Store trcs and a respository dump (rambdb_exp_dmp.zip) would probably tell us exactly what happened.


Regards,
Séamus Coffey
EMC Global Services

hi all

i just deleted the host container without clearing wwn's. infact i did delete couple of other servers and re-added them.. but the issue was on one server only.

if it affects the zoning database.. is any log available that can be used to investigate this problem?

i did raise a case with emc# 25419872 but the emc engineer was not able to analyse as the logs were overwritten...

can anyone tell me any other way to analyse the cause?

thankss..

Are you sure the logs that Seamus mentioned are gone ? In a lot of cases you only need the symapi logs, but Seamus is talking about other logs instead

i uploaded all the logs in the following site.

ftp://ftp.emc.com/incoming/25419872

not sure if u guys have access to it. U can check..

Thanks