NimbostratusHi,
We have recently set up the LTM for windows terminal server load balancing. We have 2 RDP servers in the pool. The health monitor we have set up is a TCP monitor with interval and timeouts set to 30 and 91 however we are having constant disconnects with no real pattern. Any ideas as to why this might be happening? Should we increase the interval and timeouts further? Thanks in advance.
CirrocumulusUCIRL,
You can increase the interval/timeout but you're probably just masking an issue, and if you did have a problem with the RDP servers further down the line then this course of action may have implications for users been sent to servers that are down but not yet reporting they are down in the BIG-IP configuration because of the high timeout value.
Are you able to do a tcpdump to see why the monitors are failing? What about if you changed the monitor to icmp for a test, do they stay up then? If so then this would point to tcp connections on the server being the issue.
Hope this helps,
N
NimbostratusHi Nathan. I suspect it might be an issue with one of the servers as I've been monitoring the logs and I think only one of the servers appears to be affected but I'm not 100% sure of that as it's very intermittent. I will try your suggestion to see if it is an issue with the server. Thanks very much for the response.