Subscribe

SMI-S 5.1 Agent errors

Greetings all,

Hoping there's someone with more CIM experience than I have who can point me in the right direction.  I'm getting the following errors on one 6220 HA pair but not another when the 5.1 agent attempts discovery:

18-Mar-2014 13:22:25.876 SEVERE  cimserver: CMPIError raised by provider:At ..\..\cmpiutils\filerdataimpl.c:753 in getFilerByName(): ONTAP_FilerData Instance not found

18-Mar-2014 13:42:12.778 SEVERE  cimserver: CMPIError raised by provider:At ..\..\cmpiutils\filerdataimpl.c:753 in getFilerByName(): ONTAP_FilerData Instance not found

18-Mar-2014 13:46:33.649 SEVERE  cimserver: ONTAPStatus:At ..\..\wbemontapi\utils.c:237 in setErrorONTAPI(): Filer return: Not enough memory to get instances for volume.Use perf-object-get-instances-iter-* calls and/or ask for specific counters to limit the amount of data retrieved - With storage system n.n.n.n.

18-Mar-2014 13:46:39.711 SEVERE  cimserver: ONTAPStatus:At ..\..\wbemontapi\utils.c:237 in setErrorONTAPI(): Filer return: Not enough memory to get instances for volume.Use perf-object-get-instances-iter-* calls and/or ask for specific counters to limit the amount of data retrieved - With storage system n.n.n.n.

18-Mar-2014 14:12:24.978 SEVERE  cimserver: ONTAPStatus:At ..\..\wbemontapi\devicecacher.c:750 in tellPegasusCacheReady(): Filer Cache: Could not open file mapping object

This seems odd to me as the HA pair that's working is (generally speaking) busier than the HA pair that isn't.  So if I were going to get "out of resource" errors I'd expect them to be reversed.  Things I've tried:

1. Switched from https to http filer access

2. Switched from service account to root for authentication

3. Restarted the cimserver

There has to be a config difference between the pairs that I'm missing.  Either that or both my controllers on one site have spent all day being short on RAM while the other, more I/O intensive site is using RAM more efficiently.

Re: SMI-S 5.1 Agent errors

Opened a support case and it's 5.1 making an old API call.  They're opening a new bug report and escalating to SMI-S engineering.  Hoping that this fixes it.