All questions related to installations, configurations and maintenance of Advanced Host Monitor (including additional tools such as RMA for Windows, RMA Manager, Web Servie, RCC).
I am having a problem getting Host Monitor 4.45 to monitor all of my Microsoft servers. I am currently monitoring 10 servers and plan to add an additional 80+ to the list. Some serves come back as “unknown”. The only test/events that comeback as unknown are related to PerfCounter. Ping and DNS tests are always successful. I have tried all four options related to PerfCounter checking. Anyone have any ideas?
Performance Counter test returns "unknown" status from time to time? Or all the time?
What account do you use to start HostMonitor? Performance Counter test requires administrator's priveleges.
Both. I have seen a working server change to unknown, but primarily after I restart HM. Sometime it will change from an unknown status to good. I am using a domain admin account to test the serves. It is not specific to a domain, because I have good and unknown servers in the same domain.
The source and destination are both Windows 2000 servers. The source server is a file in print server with utilization no more than 3 to 5 percent at any give time. I would even say the average is less than 1 percent utilized.
Is there a log file that details why its not connecting?
No, there are no logs
I think you can use perfobj.exe command line utility (it is included into HM 4.45 archive). Start it when test fails using command line perfobj.exe PerformanceCounter [-n retries] [login password [domain]]
E.g. perfobject "\\Server1\TCP\Connections Established" admin pswd25
It will display counter value or error description
I have something similar going on. We are running AHM on a Win2K SP4 server with 2GB RAM and dual 1.8GHz HT Xeon processors. When I test ther perfcounter objects individually or from the command line, I get results quickly. When I let AHM go through the test cycle, I end up with many counters in an unnown stat. We are running about 3600 tests. I have tried changing the tests/sec at several values from 32/sec to 100/sec.
I have something similar going on. We are running AHM on a Win2K SP4 server with 2GB RAM and dual 1.8GHz HT Xeon processors. When I test ther perfcounter objects individually or from the command line, I get results quickly. When I let AHM go through the test cycle, I end up with many counters in an unnown stat. We are running about 3600 tests. I have tried changing the tests/sec at several values from 32/sec to 100/sec.
That also works! We ran into a situation, however, where we were running so many tests on the LAN itself that the reduction in threads had a significant impact. We wanted to avoid installing RMA's on every server (about 200). We are, however, running an RMA at each remote location.
