I recently encountered an issue in Azure Operational Insights (OpInsights). Data aggregation appeared to be failing for the OpInsights capacity planning intelligence pack when only hours before, data aggregation was working as expected. I couldn’t help but wonder what could have severed the connection. In my experience with OpInsights, data aggregation failures usually indicate a problem with OpsMgr sending data to OpInsights Analytics. I didn’t know of any configuration changes had been made on the OpsMgr side, but decided to double check everything.
While scouring OpsMgr, I noticed some 2127 event IDs in the Operations Manager log on the management server, with message “health service HTTP module in management group …failed to POST data to/from url (https://ods.systemcenteradvisor.com…)” This event ID could be caused by any number of issues, and merely confirms what I’d already suspected, but doesn’t tell me much else.
So using the Performance Monitor tool, I sought to determine if data was being sent up to the System Center Advisor service. To do this, I filtered for counters that start with ‘HTTP’ in the ‘Health Service Management Groups
And sure enough I determined that there was no performance activity for the counters in question. Indicating a configuration problem. The next logical step was to review the OpsMgr-to-Virtual Machine Manager (VMM) integration, which is critical to the functionality of OpInsight’s capacity planning intelligence pack
And…voila! There was a connection error.
It appeared that the installed integration management packs were no longer supported. I subsequently determined that the VMM server which was running VMM 2012 R2, had recently been updated to UR3. This update clearly upped the supported connector and management pack versions for VMM, making it incompatible with the connectors and packs in OpsMgr. To address this, I reconfigured the VMM-to-OpsMgr connection, and confirmed that necessary performance counters were available
After reconfiguring the VMM-to-OpsMgr connection, the necessary performance counters are now available
…And after aggregating the collected data over a couple of hours, the OpInsights Capacity Planning intelligence packs is now working as expected.
This exercise prompted me to wonder how much easier and faster pinpointing the operational issues, and remediating them would be if I could tell at a glance in OpInsights, what configuration changes, if any, had occurred in my environment. Gladly, Microsoft’s OpInsights Team has come up with an answer to this, in the form of a new Change Tracking Intelligence pack in System Center Advisor. Read about it on their blog, and give them some feedback, so that they can make well informed decisions on what additional change tracking scenarios to enable.