WHILE UPGRADING TO HIGH AVAILABILITY MODE THE PRIMARY AND SECONDARY SERVER SHOULD BE ON SAME SUBNET(BUT PRIMARY AND SECONDARY SERVERS ARE IN DIFFERENT LOCATIONS)

I recently had enabled a SolarWinds HA pool for our primary server and had a boat load of questions. I opened a case with support around expected functionality and they pointed out a couple links.

Here's one specifically towards your scenario. This setup that support pointed me to I thought was intriguing. Not sure if SolarWinds supports it though but...

https://thwack.solarwinds.com/thread/109227

Also here is a good read for SolarWinds HA...

High Availability in NPM - SolarWinds Worldwide, LLC. Help and Support

Failing over in my environment takes approximately 1-2 min. Note web console sessions will expire and all users will need to log back in as necessary.

Hope this helps!

giais

Thanks for the response Giais. I am unable to understand the Process. We are implementing the Orion Servers on esxi. Can you explain How I can choose Ip's for primary and secondary in the same subnet?

From what I understood from the post, is that each SolarWinds server that is going to be part of the pool is given an additional network interface with an IP subnet of your choosing. This subnet would not route, or have connectivity, and the 'shared' IP is also in that shared non-routed subnet. It's simply to get around the requirement that the servers in the HA pool need to be in the same subnet.

If I were going to implement something like that (which I'm not), my HA setup is all in the same subnet -- I'd work with SolarWinds support directly to see if it's supported, recommended, etc..

You need to create a virtual IP so that the failover is seemless.

What is a Virtual IP address in NPM? - SolarWinds Worldwide, LLC. Help and Support

I've done what you're trying--two HA servers, different data centers.  I implemented Cisco's NX-OS and leveraged their Overlay Transport Virtualization--OTV.  It enables you to extend a Layer 2 VLAN / Subnet between two different sites, while retaining the ability to block broadcast storms between the two data centers.  We ended up with many HA servers on the same IP network but in two different locations through using OTV.

Now that's pretty cool.

But in my environment, we are not using nexus 7k .Is there any other way to connect with catalyst environment?.Thanks in advance.

teja0909​, do you own High Availability licenses or are you evaluating HA in your environment?

aLTeReGo Yes we have HA licenses.

To the best of my knowledge, no, there is not a different way to extend an L2 subnet/VLAN between two physically separate/remote sites that meets HA criteria.  Although you should work with your data center / WAN architect and VAR to confirm you can't come up with something elegant.  You shouldn't kluge together a one-off that isn't robust and doesn't follow network design best practices.

I think most of us could build a kluge that might accomplish this--maybe you'd create a GRE tunnel between the sites, maybe you'd dedicate a Layer 2 Interface at both sites and physically span your L2 between them--certainly not a best practice!

Work with your VAR and WAN hardware manufacture's Sales Engineer and technical support staff to ensure your end result is something that will do the job right for you and your company for years to come.