numerous strange issue with Orion 9.1 SP5

I'm not sure where to start, but I am seeing several strange (interrelated?) issues with my installation of Orion 9.1 SP5 on four different servers.

I have a primary, hot standby, additional web server and DB server. Each server is chocked full of memory, fast processors and far exceeds that minimum specs listed by SW in their documentation. The current monitored network is 4179 elements.

Here are the issues I am seeing:

(1) The primary poller frequently loses communication with the DB server. Right now, it is showing the last DB sync was 19 hours ago. Most of the nodes on the web site show down, which I know is not correct. the SW services are all up, but when I try to launch the System Manager and connect to the primary poller, I get the following error message (see attached screen shot).

(2) I was able to effect a failover to my standby poller, but the standby goes up and down every few minutes (engine status shows stopped). I have attached a couple of screenshots of the Monitor Polling Engines app. Both the primary and standby pollers engine status alternate between stopped and running states. Is this normal?

The System Manager on the standby poller shows no elements at all, even though the website shows the same as the primary poller website (many nodes being down).

non-connection.JPG

Find more posts tagged with

web_console

primary_poller

system_manager

additional_web_server

Failover

Accepted answers

All comments

freemen

The second screen shot:

StandbyEngine-Green.JPG

freemen

and the last screen shot:

StandbyEngine-Red.JPG

Yann

Hi,

For the compatibility error follow the steps of this KB Article: http://support.solarwinds.com/kbase/ProblemDetail.cfm?ID=938

Use these steps to resolve this issue.

1. Stop all Orion services
2. Navigate to Start->All Programs->SolarWinds Orion->Advanced Features and launch the application called Monitor Polling Engines. You may get informational messages but this is OK as long as your services are already stopped.
3. From the toolbar across the top of the screen, find the icon that is a picture of a silver box next to a red box, the tooltip says Poller Load Balancing. Click this icon. You may get one or more informational messages.
4. Select the top node in the list. Scroll to the bottom. Hold down shift and select the last node in the list. This will select all of your nodes.
5. Find the Polling Engines pull-down menu (next to the File pull-down menu at the very top of the window). Select it choose "Move selected Nodes to" then select the name of your polling engine. Even if you are assigning all your nodes to the same polling engine, do this anyway.
6. Click Done, close all windows and open the System Manager.

If these steps do not resolve your situation please create an online ticket to ask for help from Technical Support.

NOTICE: If you did not stop your services before you made the change, then your changes will be undone when you click the Done button. Be sure your services are stopped before you reassign your nodes.

Concerning the status of the polling engines in the tool, ensure the time is synchronized between the servers (including the Database server).

HTH,

Yann

freemen

Thanks Yann. I understand the steps you have outlined, but my other poller is a hot standby, not an additional secondary polling engine. Does that matter?

Yann

Ensure the services are stopped on both servers (Disable them if needed) for the compatibility steps.

if the time is not sync between all the servers (db included) it could explain why the Hot-StandBy fires up all the time.