Forum Discussion
NimbostratusCheck VIP availability from a different VIP Pool
Hi Folks, My App team ask me to host their application at LTM and GTM. I am stuck at a point for a condition.
Scenario: Application is stage up at three layer. Please refer the image for the application traffic flow.
** vs-1 hosted app is dependent on the availability of vs-2 app. ** Traffic needs to pass vs-1 and then vs-2 aap.
Condition: ** Need to monitor the traffic reaching at VS-1 would be able to reach to last node of vs-2 pool member. ** would like to have a health monitor at pool-1 for monitoring vs-2 availability + pool-1 member health check ** Can I achieve this at VS-1 or poo-1 or, by any chance?
Thanks in advance!
/Nantu!
Soda_Cup_148395if you want to monitor something it needs to be in a pool. if you place monitors on both the pools you have identified I think you will find that the functionality you require is already in place.
1)VS-1 --> will only send to pool 1 if available members. 2)pool members use VS-2 as a resource(in your app) and VS-2 will not forward requests to pool 2 unless there are available members.
Or if you only want traffic forwarded past VS-1 ONLY IF pool 2 has available members, use an irule to check pool 2 health when traffic first hits VS-1.
NANTU_271091Hi, your second assumption is correct. I want to pass the traffic at VS-1 if VS-2 is available. Now tell me one thing.
Can I use two monitor profile at Pool-1 (one is default AND another monitor for monitoring VS-2 and port. So, both the monitor value should be 1, if anyone is 0 VS-1 would be down. So that WIDE IP top of VS-1 will not forward the traffic to VS-1.
Also what would be the iRule should I use what you mentioned.
/Nantu
DevBabuCirrus
For your current example.
For Pool 1, You need to create three monitors.
- Default Pool monitor that will monitor node-1 and node-2.
- Monitor to monitor node-3 and it's port.
- Monitor to monitor node-4 and it's port.
2 and 3 can be achieved by creating Alias address and Alias port in the monitor. Alias address will contain address and port of node-3 & node-4. Other request response string same as for monitor created for pool2.
Now apply that monitor 1 i.e default to the pool.
Apply monitor 2 & 3 to the pool member (click node1:port, Configuration Advanced, Health Monitors- Member Specific, Select monitors 2 & 3, Availability requirement atleast 1)
Repeat for node2:port
So, if both monitors 2 & 3 fail i.e node 3 and 4 fail then Pool member will fail and the VS 1 should go down.
I have not tested this but hope this works. Let me know your result
- Soda_Cup_148395
It would be easiest to configure this through iRule, there are a lot of resources to write this and test it yourself, but here is an idea-
when HTTP_REQUEST {
if { (active_members POOL-2 == 0) { reject or http redirect here }}
This would be applied to VS-1 where POOL-1 is the default pool. At each http request to VS-1 the status of POOL-2 would be checked before sending the request to POOL-1.