How do I restart the DPA service without having the CMD window continue to stay open? I've found the "startup.bat" file. Do I have an option of running it as a windows service perhaps?
I believe I've found an answer to my own question. I ran the "IgnitePI_StartServiceX.vbs" in the Confio/Ignite PI folder and it seem to turn the service without showing the CMD prompt.
Awesome - let us know if you run into any issues though or reach out to support at Submit a Customer Service or Technical Support Ticket - SolarWinds
This looks like an older thread but, I wanted to clarify it some as It could see this one causing some confusion. When DPA is installed on a windows server it should create a service out of the box. The service is called "Ignite PI Server"
As the name has never been changed from the Confio days of DPA being Confio Ignite.
If the service doesn't exist the fast way to restore it is a .bat file that is in the root folder of DPA or default path on the current release is C:\Program Files\Solarwinds\DPA\ called CreateWindowsService.bat on older installs it might be found in C:\Program files(x86)\Confio\Igntie PI. This does call some .VBS files to check for java versions and then build the service.
The service is needed to be recreated if Java Heap memory max settings are changed to scale DPA for the number of monitored instances so we in use this .bat file and the removeWindowsService.bat file often for that task.
Steps can be found here.
Increase DPA heap memory to monitor large numbers of instances - SolarWinds Worldwide, LLC. Help and Support
We had an issue with a monitored server rebooting and DPA giving up on re-establishing monitoring after a period of time. We don't want to manually intervene. Is our best option to shutdown and restart the Ignite PI Server (I assume DPA will re-establish monitoring with all configured databases)? We could do that via command line using our scheduler (TIDAL).
Eric, how long was the monitored instance down? DPA should continue to attempt connections to the monitored instance, with less frequency over time. Can you describe the exact sequence of events that happened when DPA did not automatically connect again to the instance?
Restarting the entire DPA Server seems like overkill but could be an option. Your scheduler could also check the status of the instance in DPA and start the monitor if it finds one down:
select id, name, status from cond; -- status could be STOPPED, STARTING, STARTED
-- if you find one of them STOPPED when it shouldn't be, it can be triggered to start
update cond set command = 'START' where id = <id from above query>; -- command can be STOP or START
