This is an old revision of the document!
Table of Contents
Grid Master Candidates
Marking a Grid Member as a GMC will not cause a reboot but it will cause services restart on that specific Grid Member alone because it needs to receive a complete binary database from the Grid Master and apply it rather than the partial XML database it currently has. The database switch-over involves an automated restart.
Promoting a Grid Master Candidate to Grid Master will restart service on every device in the Grid (but the devices will not reboot). There is an option to do this sequentially but you cannot controll the order. The GMC being promoted will restart first, then the old GM, then every other Grid Member in FQDN order.
Test Promote Grid Master Candidate
In the GM GUI, you can run GMC Promote Test where you can select all non GM/GMC nodes and test their communication to a GMC of your choice.
Connectivity to Port 1194 and Port 2114 between the non GM/GMC members and the GMC you selected will then be tested and the results made available.
Promote Grid Master Candidate
set promote_master
You can do one of the following to promote a Master Candidate:
- Immediately notify all Grid members about the promotion.
- Set a sequential notification to provide wait time for Grid members to join the new Grid Master. Staggering the restarts of Grid members can minimize DNS outages. The sequential order for Grid members to join the new Grid Master begins with the old Grid Master and then the Grid members in FQDN order. The default delay time is 120 seconds. You can configure the delay time from a minimum of 30 seconds up to 600 seconds (10 minutes).
In my lab, the TE-815 NIOS DNS servers took 3.5 minutes (ish) to reboot and rejoin the new Grid Master. For this reason, I would normally put 4 or 5 minutes as the delay time.
During a promotion, the members will connect to the new GM, re-sync and perform a services restart (not a product restart and no reboot)
GMC Error
If you have a VM Grid and the GM says that the GMC is not suitable for GMC role, it could be that the GMC hard drive is too small. I had this when I deployed a NIOS 7.3.11 VM that had a 55Gb hard drive. This image is really for Multi-GM only. GM and GMC should use the 250Gb image.