IMC HA Learning

[WillH]

HI, Does anyone know where I can go to learn about IMC HA? The only thing I can find is an install document but that doesn't really tell me much about what it actually is and how it works. Thanks

[bala5]

Hi,

There are 2 types of HA available.

1. Stateful.

https://support.hpe.com/hpsc/doc/public/display?sp4ts.oid=4176520&docLocale=en_US&docId=emr_na-c04958702

Needs licesnse to be installed.

2.Stateless.

Refer the below guide 

https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-c05162576

for the second scenario, the backup server will be take over the master role but you can t add or edit any exisiting config.

Only old master allows you to edit the config.

Thanks,

Bala

[NeilR]

There is the "warm" standby server option which is similar to stateless. Does not require SQL server if you'd rather not go that route and just uses mySQL on each.

Create a second server and install the same license which will create a back up server - same restrictions - no editing just monitoring (except in UAM)

Set the nightly backup to automaticaly restore on the standby node. This will create a daily copy. 

Set all the traps, syslog etc to point to both servers. The standby will alert and alarm like the active one. You can recover all the alarms at once (but not individually)

If you are using UAM for 802 or other authentication, make the second server the second authentication source on all your devices. UAM is totally active - you can synch users, modify policies and services etc. Most devices will fail over to the 2nd server if the first one times out.

Anything that was configured before the nightly backup will be included. So any nodes, performance management tasks etc will come over.

Advantages of this approach: simple to configure and ability to separate operations, aside from daily backup, which allows upgrades and testing independantly. The other server takes over critical functions while the primary is down or being worked on.

As we run IMC on VMs, for an upgrade I can snapshot both, upgrade and test the primary, while the backup handles the basic functions of authentication and monitoring. If it works ok I can then upgrade the standby and turn on the backups again. If not I can roll back to the snapshot on the primary.

If your enviornment is very volatile or you can't lose any operations or performance data then this mode is not the best for you.