Forum Discussion
CirrusHealth Monitors
Hi everyone I have 2 ltm_https pool members. i used gateway_icmp health monitor for the pool and none for the nodes however the pool and virtual servers became available while the nodes are unknown enabled. what will be the reason and which health monitor is best suited for https server? Regards
RaghavendraSYAltostratus
Hi,
You can assign monitors at three levels: 1. Pool 2. Pool Member 3. Node
Please verify what monitors are assigned at each level. Best option at pool member level is "inherit from pool".
If you configure gateway-icmp as monitor, pool members should be available if you are able to ping servers.
As i said please verify monitor at node level too and let me know the status
Lee_SutcliffeNacreous
Health monitors filter up from the node, to the pool to the vip. So if you do not have a monitor at the node level, their status will not be passed to the pool level monitor. This is why your VIP and pool are showing 'green'
Good practice is to enable ICMP monitoring at a node level and have a more specific monitor at the pool level - such as an HTTPS monitor.
If ICMP fails at the node level, it will filter up to the pool level monitor and mark it down, which in turn will mark down the VIP
- Kevin_Davies
Use a https monitor on the pool. You don't need a node monitor as the pool monitor implies the node is up if it is working.
The login page is usually a good option. Choose some text on that and note the URL. For example if the text was "welcome to corporate site" and the url was https://www.site.com/index.html then in your https monitor the send string would be...
GET /index.html\r\nand your receive string is...
welcome to corporate site
