IMC
cancel
Showing results for 
Search instead for 
Did you mean: 

Distributed IMC implementation options

David-G
Occasional Visitor

Distributed IMC implementation options

Hi all,

I've been through a single IMC Enterprise setup and config and all went ok as expected. It's a great platform with lots of useful functionalities able to cover a multivendor network.
Right now I'm trying to test a distributed IMC implementation where I would like to have two management servers for the same distributed network, for redundancy purposes, as if a server is unavailable at one site, I can always use the other.
I've been reading through the documentation, and there seems to be available only a limited deployment mode, where the Master contains some modules and the Slaves contains some another modules, that can't be present in the Master. It's kind of a deception that I can't seem to run the same module on more than one server, as I was able to verify on my tests.

So what are my options? Should I just implement two independent servers managing the same end devices? What about devices and config template consistency across two platforms?
Any help/recommendation/design tip is welcomed, as I'm probably missing something here... I was expecting more from a "Highly flexible, scalable deployment models" platform.


Thanks in advance,

David Galveias

4 REPLIES
pombeii
Frequent Advisor

Re: Distributed IMC implementation options

Hi

I think you confuse Distributed with HA. For HA, you build 2 iMC servers so when one server has problems, you can use another.
For Distributed, you use several servers to deploy different IMC modules, but all servers work as a single IMC system.

mbrito
Occasional Visitor

Re: Distributed IMC implementation options

Hi Guys!

 

I have a similar question. I manage a network with about 150 devices (switches and APs) and I run IMC Std. 5.2 on a Linux box with MySql. How can I implement IMC redundancy ? Would I need another server with IMC and a new data base? Or they can share the same Mysql data base?

 

Regards

 

Peter_Debruyne
Honored Contributor

Re: Distributed IMC implementation options

That would be 2 totally separte installs (with separate DB, since you want failover including a possible DB failure).

Base principle is to use the backup on the primary server to a shared path, and run an automated(scheduled) restore on the secondary. This will keep the DBs "kind of" in sync.

All devices should be configured to send traps to both IPs.

 

Hope this helps,Peter.

C0LDWiR3D
Frequent Advisor

Re: Distributed IMC implementation options

So basically you fill both the Active and Standby environment with syslog and snmp traps and the backup/restore keeps the installations in sync ?