I am getting more and more charts, that are missing information - what could be the reason for this?
Hi,
The data gaps may be caused by your configured polling interval not able to cope up with the number of elements in your polling engine. You may use the Polls per Second Tuning tool to properly tune your polling engine. Just go to your server --> Programs --> SolarWinds Orion --> Advance Features. You will see there a recommended setting which is based on your configured polling intervals and number of elements. Please be aware though that resetting the values will need to restart the services.
Also, if the tuned settings are too high, this may cause a sudden rise in CPU and if it indeed it happens, you may need to upgrade your hardware or consider another polling engine to load balance the elements you are monitoring.
Hope this helps
I was having the same issue when my SQL and Orion servers were VMs.
Support could never resolve it and would typically give me 1 of 3 answers: 1: There's an issue with SQL. That "issue" was never identified. 2: The VM doesn't have enough resources 3: Known issue with the polling engine that is fixed in the latest version (got that answer for every version/SP between 10.0 and 10.1.1).
Problem finally disappeared when we migrated to dedicated large physical servers (quad-core, quad-proc 3.4Ghz Xeon, 32G RAM) for both Orion and SQL. Additional polling engines are still VMs and having no issues with NPM. We also migrated to 10.1.3 at the same time.
While I was troubleshooting I was looking at the DB tables and could see that there was no data for the time periods corresponding to the gaps. My SQL admins looked at the servers and could see no issues. Likewise, the VMware admins could see no issues but still did some NIC tuning and gave me plenty of CPUs and memory.
