In a fault tolerant environment, the event information is saved to the secondary SpectroSERVER database when the primary server fails. If the primary server had both the SpectroSERVER and Archive Manager stop, then proper procedure is to start the Archive Manager first, so that SPECTRUM will transfer the events from the secondary SpectroSERVER to the mysql database on the primary SpectroSERVER. In order to have this done automatically when the box is rebooted, you need to modify the ARCHMGR.idb file in the <SPECROOT>/lib/SDPM/partlist directory on the primary SS only.
You will need to modify the following lines in the ARCHMGR.idb file on the primary SpectroSERVER from:
Once this change has been made, you MUST restart the SPECTRUM processd for the changes to take effect.
Once this change has been made, you MUST restart the SPECTRUM processd for the changes to take effect. Keep in mind that shutting down the processd does stop the SpectroSERVER.
On Windows, you can navigate to the Windows Services and cycle the SPECTRUM Process Daemon.
On Solaris and Linux, open a shell window and navigate to the <SPECROOT>/lib/SDPM directory and run: ./processd.pl stop
To restart it, run: ./processd.pl start
In a fault tolerant environment, how can the Archive Manager on the primary SpectroSERVER be automatically started without starting the SpectroSERVER?
If the primary SS fails, or the box is shut down, powered down, or hard booted, how can the Archive Manager be configured to start automatically before the SpectroSERVER?
In order to save or to not lose the events from the secondary SpectroSERVER, how can the Archive Manager on the primary be started automatically?
Spectrum 8.1.0 Core
Keep in mind that you will need to monitor the local store of Events on the VNM model in OneClick prior to starting the primary SpectroSERVER. If you start the primary SpectroSERVER before waiting for the local store of events to reach 0, then you will lose those events.
(Legacy KB ID CNC TS30433 )