We have 2 DHCP servers with failover realtionships (hot stanby mode).
Spontaneously dhcp service hangs. The process consumes about 200 Mb of memory and 25% of CPU contantly. You cannot expand the tree in DHCP mmc - hangs.
Clients stop to receive IP's. When checking the status from the second partner it says status (!!!) normal. It doesn't failover!
We tried to find the root cause. We noticed that one scope has a noisy PC, reqesting very frequently DNS update operation. DHCP log becomes full and the logging paused. After some time passes the service hangs. Only service restart helps.
Please, don't advise to increase the log size - already done. It's not the full solution. I have a dump made from task manager of hung process, can send you by request. So, please help to push the вумеуфь to make a patch.
Thanks.