Does anyone know if\when the following features will be available?
1. iSCSI boot support for vSphere 5
2. iSCSI initiator iqn configurable by service profile template
Lack of these features is making iSCSI boot a bit of a non-starter for designs I'm working on at the moment.
I have the same problem with Jeremy Waldrop but on 6120XP
Setup with new the software 2.01q (upgrading and erase configuration)
Did some further digging & investigation, it seems if you had previously loaded beta code, this alarm will be present. If you only transitioned from release builds to release builds it will not display.
Jeremy - That show tech you uploaded shows it was previously running a beta build 239 at one time -
Vendor: Cisco Systems
That explains it then. In some of the beta builds we exposed the iqn pool code, which would have created the iqn managed object in UCSM database. Even when going to newer released code, we may mask certain objects (such as the iqn pool from being displayed), but we don't delete them. So because iqn pools weren't commented out of Beta code, it will be an artifact until you upgrade to the Maintenance Release in Jan (tentative).
Robert - I upgraded from 1.4.1i straight to 2.0.1t and I am getting the same iqn pool default is empty error.
I know this is just a cosmetic error not afecting anything in the environment.
But is there a work around to remove this from being visible in UCSM? Maybe thru the CLI?
This is with a pair of 6120s.
The issue Jeremy hit was with the IQN pool, which was what Rob was addressing.
If you have not used any Beta builds and have upgraded from Cisco.com releases (as you mentioned, 1.4.1i to 2.01t), you are most likely hitting an issue with the iSCSI IP Initiator pool being empty. Workaround is to specify some values in the pool. Unfortunately this pool cannot be deleted, even if you're not utilizing iSCSI Boot. So your options are to either ignore the "fault" or enter some dummy values into the IP pool.
There are 2 different iSCSI related faults in UCS 2.x. One is correctable the other isn't. The fault about the iSCSI IP Pool is correctable by adding a single bogus IP to the iSCSI IP Pool.
I have only seen the other fault appear on 6248 Fabric Interconnects. This fault is about the iqn pool being empty but there isn't an iqn pool exposed to UCSM.
gso-ucs-lab-A# show fault
Severity Code Last Transition Time ID Description
--------- -------- ------------------------ -------- -----------
Minor F0821 2011-11-01T17:30:46.060 44695 iqn pool default is empty
I would be nice if empty pools didn't trigger a fault.
Thank you both for your responses.
The iSCSI IP Pool was easy to identify when I did the upgrade and I applied a generic IP to it to clear the error.
The one I mentioned is for the iqn pool. Which is nice you are getting an error that a pool is empty and that pool doesn't exist or is not exposed to UCSM
and I agree with you Jeremy, empty pools shouldn't trigger a fault.
Some heads up on the iSCSI capabilities in the recent MR1 update. Though the IQN is pooled, there is no validation done to ensure it meets the naming standard. Arrays such as NetApp and EMC will validate any initiator's IQN for correct format. So right now it's a manual check process within UCSM to ensure the standard is followed.
I have opened some enhancements to to enforce validation, which will likely be added in the next patch release.
Also note that only IQN format are supported in the iscsi initiator pool resources. EUI can be statically entered, but not pulled from a pool at this stage.
For those of you that haven't upgraded to 2..0.2m and want to get rid of the IQN Pool default empty alert, here is a way to do it thru power shell scripting.
Just connect to your UCSM instance and run this line.