*
News: SMF - Just Installed!


Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length

CIMOM connection lost error when performing inventory of VMware ESXi (OS) 2572

Super Newbie Posts: 6 Karma: +0/-0 *
CIMOM connection lost error when performing inventory of VMware ESXi (OS)
« on: February 09, 2015, 11:16:57 AM »
Hello everybody,

When performing inventory of any of my VMware ESXi nodes (resource type: Operating System) I receive following error/warning message in 'Collect inventory' job log of my Flex System Manager: "ATKSRV630E Managed resource "OS_RESOURCE_NAME_HERE" lost the CIMOM connection. The reported inventory data might be incomplete. Verify the connection between the management server and the managed resource, and verify that CIMOM is running on the managed resource. Next, collect the inventory again.". I have verified the connection between FSM and ESXi node's OS (Actions -> Security -> Verify Connection) and it is OK. FSM accesses the ESXi server's OS using root user credentials and the Access state is OK. I'm running Flex System components on release level 1.3.1.2. What could be the reson of the error message, what are its consequesces and how to fix it?

Cheers,
Konrad
Administrator Sr. Member Posts: 309 Karma: +6/-0 *****
Re: CIMOM connection lost error when performing inventory of VMware ESXi (OS)
« Reply #1 on: February 09, 2015, 12:55:33 PM »
Hi Konrad,

Welcome to mypuresupport!

Concerning your question. You shouldn't need to inventory the OS MEP (managed end point) for ESXi (we get the same message in our lab). An inventory of the Server MEP will give you everything you need. If you want inventory down to your VM level you would discover and manage your vCenter server.

Let me know if that makes sense.

Thanks!
Super Newbie Posts: 6 Karma: +0/-0 *
Re: CIMOM connection lost error when performing inventory of VMware ESXi (OS)
« Reply #2 on: February 10, 2015, 05:09:33 AM »
Hi again,

Thank you for quick answer. Yes, it makes much sense, however there is one thing which still worries me, when it comes to inventory. I performed inventory of the server MEP of the same ESXi node and the result is the same as for OS MEP (event ATKSRV630E). See full job log below:

February 9, 2015 5:09:24 PM CET-Level:1-MEID:0--MSG: Job "Collect Inventory - February 9, 2015 5:09:21 PM CET" activated.
February 9, 2015 5:09:24 PM CET-Level:200-MEID:0--MSG: Subtask "Collect Inventory" activated.
February 9, 2015 5:09:24 PM CET-Level:200-MEID:0--MSG: Starting clients
February 9, 2015 5:09:24 PM CET-Level:100-MEID:0--MSG: Clients started for task "Collect Inventory"
February 9, 2015 5:09:24 PM CET-Level:200-MEID:0--MSG: Subtask activation status changed to "Active".
February 9, 2015 5:09:24 PM CET-Level:150-MEID:0--MSG: ATKSRV629I Collecting inventory by using inventory profile "All Inventory".
February 9, 2015 5:09:24 PM CET-Level:150-MEID:12595--MSG: ATKSRV628I Collecting inventory for "NODENAME SRV", which has a type of "Server".
February 9, 2015 5:09:24 PM CET-Level:1-MEID:0--MSG: Job activation status changed to "Active".
February 9, 2015 5:09:24 PM CET-Level:200-MEID:0--MSG: Subtask activation status changed to "Active".
February 9, 2015 5:09:24 PM CET-Level:100-MEID:12595--MSG: NODENAME SRV client job status changed to "Active".
February 9, 2015 5:19:28 PM CET-Level:150-MEID:12595--MSG: ATKSRV630E Managed resource "NODENAME SRV" lost the CIMOM connection. The reported inventory data might be incomplete. Verify the connection between the management server and the managed resource, and verify that CIMOM is running on the managed resource. Next, collect the inventory again.
February 9, 2015 5:19:28 PM CET-Level:100-MEID:12595--MSG: NODENAME SRV client job status changed to "Complete".
February 9, 2015 5:19:28 PM CET-Level:200-MEID:0--MSG: Subtask activation status changed to "Complete".
February 9, 2015 5:19:28 PM CET-Level:1-MEID:0--MSG: Job activation status changed to "Complete".

Where could be the problem?
Administrator Sr. Member Posts: 309 Karma: +6/-0 *****
Re: CIMOM connection lost error when performing inventory of VMware ESXi (OS)
« Reply #3 on: February 10, 2015, 03:51:43 PM »
Hi Konrad,

Is CIM enabled and running on the ESXi host itself? Also a good question I forgot to ask is are you running the IBM Customized ESXi version supplied by IBM?
Super Newbie Posts: 6 Karma: +0/-0 *
Re: CIMOM connection lost error when performing inventory of VMware ESXi (OS)
« Reply #4 on: February 13, 2015, 10:11:10 AM »
Hello,

I believe that CIM is enabled on each of ESXi hosts. At least the output of the command '/etc/init.d/sfcbd-watchdog status" is "running' (based on this article: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1025757), all CIM providers are also enabled on ESXi hosts (as per this website: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2008776). During last PureFlex components update I have also applied component called 'IBM Customization Patch 8' so I believe all my ESXi hosts are "IBM-compatible".

Today I restarted System Management Processors of all nodes with ATKSRV630E inventory error and out of 6 'Server' endpoints with inventory errors - 2 went through the inventory without error after the IMM restart. I have also restarted all the management agents on all ESXi hosts with this problem (using 'services.sh restart' command). To summarize, problem is still not fixed for me and FSM behaviour seems to be 'random'. For sure I will restart everything: ESXi hosts, IMM modules, CMMs and FSM before next PureFlex components update, but even if this works it still doesn't sound like a solution...
Administrator Sr. Member Posts: 309 Karma: +6/-0 *****
Re: CIMOM connection lost error when performing inventory of VMware ESXi (OS)
« Reply #5 on: February 13, 2015, 02:53:54 PM »
Konrad,

Thank you for the detailed information. In short the FSM can be flakey at times and to be complete honest I do not know why it does what it does sometimes. A good thing to try on one of those servers having an issue would be to completely remove it from the FSM and rediscover it. This has cleaned up weird issues in the past. If the issues persist AFTER a removal and rediscovery then I would suggest pulling in IBM support.

Keep us posted!