Forum Discussion
Kleython_Kell_5
Nimbostratus
NimbostratusCustomize Monitor for link
Hello people, good morning
Now, I Have 2 Links, LINK A and LINK B, and i am using
teh default monitor bigip_link, but i need a monitor more eficcient. In some cases the link is up, but not operational. The I need to create a customize monitor
using an irule, i need this irule do a ping test in two or three trusted sites and verify the response time, and with this information, mark the link down if
the response time of the 3 sites were more than X.
I am using version 9.4.8
Is there some irule to do something like these?
thanks a lot
Att
Kleython Kell
6 Replies
Kleython_Kell_5Does anyone have any idea how to make a monitor with these features? Or a better idea of links to monitor, considering when the link is dead but does not fall?- Kleython_Kell_5Usint transparent monitor, i can just icpm test in anothers ips, but is there any way to use response time in the test? thanks
ArieAltostratus
Could you clarify what you mean by "the node is up, but not operational"?
It sounds like you're saying that the node responds to ICMP, but that the service running on the server is not responding. If that is the case you can probably resolve the situation by adding a different type of monitor (e.g. HTTP).
- Kleython_Kell_5Its a monitor for a LINK , in Link controler, using bigip_monitor. When i said is not down but not operational, happens when the link is slowly, with 5% of your capacity for example, but up. Then a monitor with icpm test but with response time could be a solution. any idea?
- Kleython_Kell_5its not a server,, its a link, Its a monitor for a LINK , in Link controler, using bigip_monitor. When i said is not down but not operational, happens when the link is slowly, with 5% of your capacity for example, but up. Then a monitor with icpm test but with response time could be a solution. any idea?
- Arie
You could probably write a custom (external) monitor. Another option would be to tap into another monitoring solution (e.g. Nagios, or even a custom server-side health probe) and use a regular monitor on the LTM to query it.
