StandAlone Orion Platform to HA Orion Platform

Dear All,

We are currently working on a new Large SolarWinds deployment, Due to Infrastructure availability company is currently suggesting to implement stand alone Orion Platform at first and then upgrade to High availability

Below are the expected components ( NPM,NTA,NCM,IPAM)

- Orion Management server

-APE

- Orion Database

- NTA FSDB -SQL

Orion management server and APE High availability will be covered by SolarWinds HA features, So will I face any issues or down time while creating a new HA POOL ???

Q: N/W device will be monitored already by the Stand alone APE, If I add the stand alone APE into HA Pool, what will happens to existing polling , Do I need to define the monitoring again for the N/W devices using the newly defined HA POOL??

At the network device End.

I understand,, that Netflow destination should be updated with new virtual Host name or Both APE Pool Member's IP.

Looking forwards for some pointers and help to understand in advance the road block ahead.

Find more posts tagged with

npm

Accepted answers

All comments

dgsmith80

Hi dhinagar_j

dhinagar_j wrote:
Orion management server and APE High availability will be covered by SolarWinds HA features, So will I face any issues or down time while creating a new HA POOL ???

A: No - You will deploy the HA Software onto the second machine, and once it is operational you will then create an HA Pool. As long as the current server is going to remain the Active Primary then there will be no downtime. The only other time you might need downtime is if you need to re-IP the current server to use that IP as a VIP.

dhinagar_j wrote:
Q: N/W device will be monitored already by the Stand alone APE, If I add the stand alone APE into HA Pool, what will happens to existing polling , Do I need to define the monitoring again for the N/W

A: Any Nodes set to the current APE will automatically be assigned to the active server in the HA Pool. The Nodes don't show the HA Pool in Polling Engine it shows the currently active member of the Pool. Example SOLARAPE01 & SOLARAPEHA01 in a Pool, if SOLARAPE01 is the Active Primary that will be what is shown in the Polling Engine part of All Nodes. If you fail over to the SOLARAPEHA01 then that name will be shown. Hope that makes sense. But no downtime or additional work to do once you add HA.

dhinagar_j wrote:
At the network device End.
I understand,, that Netflow destination should be updated with new virtual Host name or Both APE Pool Member's IP.

A: If I were you I would suggest creating a DNS Name for your Traps, Syslog & Netflow to point to, which in turn points to the HA VIP Pool. This makes it easier to change later on down the line. If you're going to scale you might want to consider putting a load balancer with some Kiwi Syslog Servers to reduce the load on your Orion DB.

You also have the ability to move your NTA Flow Data into SQL 2016 SP1 which will allow you to look at Cluster or AlwaysOn for DR/HA. You can cluster the NTA Flow Store as it is, but with the new release now GA I would suggest it might be better to move to NTA4.4 and use SQL.

Hope that helps.

- David Smith

Kenson - SolarWinds Authorized Partner

Training, Licenses & Professional Services

- LinkedIn

- Twitter

- Facebook

dhinagar_j

Dear David,

Sincere Thanks for your valuable comments w.r.t to my queries, It helps a lot.

RichardLetts

while a DNSname pointing to the HA pool address works, what also works is having your network devices send the syslog and trap devices to the IP addresses of the two polling engines in the HA pair. The syslog and trap processes are only running on one server at a time, so you do not end up with duplicated data in the database.