PCNS 3.0.0 takes 10 minutes for console to start

[Anonymous user]

I'm a little frustrated that APC support wasn't very helpful in this issue, so I'm posting it here in hopes someone else has run into the same issue.

The problem is that when the PCNS service is started, it takes over 10 minutes before the console is available via HTTP or HTTPS. There are no apparent errors in the Windows event logs, and the taskmanager shows very little CPU use. For 10 minutes, there just appears to be nothing happening, and netstat shows that the listening ports for 3052 and 6547 have not been opened. Then after that 10 minutes, suddenly the ports are listening and the console is available.

I'm frustrated trying to figure out what is going on with this, and APC support didn't really understand the problem at all. I'd appreciate if anyone can provide some suggestions.

Environment:
Server 2008 R2 Standard
JRE 6_25 (manually installed)
PCNS 3.0.0
SMX3000RMLV2UNC

Thanks in advance!

[Anonymous user]

I think I just identified the problem....

The server has a dual Gb NIC, but only one is connected and in use. The other interface is still enabled, even though it's not connected. I disabled the unused interface in Windows and the PCNS console immediately came up. I restarted it again to be sure, and again the startup was immediate. I also re-ran the configuration wizard, which then registered the proper IP address for the NIC interface, rather than the default 127.0.0.1 for locahost.

It would appear as though PCNS was completely thrown off by the one disconnected NIC interface, even though the other interface was up and running just fine.

Thanks for your help! Also, if you have any insights into why this was happening, I'm sure it would be helpful to others down the road, or to APC for correcting that in future releases.

Message was edited by: cosadmin

[BillP]

ah ok well ill pass it along like you said incase it comes up. thanks for reporting back with your fix.

[BillP]

just to clarify - in this 10 minute time period, does the service actually indicate its started or does the progress bar show its still trying to start the service? from what you say it seems like the service starts within 10-15 seconds as it should and then indicates Started but is not accessible via the web browser. are you checking this on the server itself or remotely via IP address on a different computer? what does the event log in PCNS say if you access PCNS when it is running after encountering this issue?

[Anonymous user]

Yes, the service IS fully started within the 10-15 seconds, but he console is unavailable. I'm accessing the console both locally, and remotely. Neither work. I've used netstat to confirm that the ports (3052, 6547) have not been opened.

When I can finally access PCNS, there was a log entry about not being able to communicate with the NMC, but I was able to correct that. Interestingly, when I re-ran the configuration wizard, it registered 127.0.0.1 with the NMC instead of it's actual IP address.

Also, this was an upgrade from PCNS 2.2.4 (which works fine). I've tried completely uninstalling PCNS and deleting it's directories from Program Files and Program Files (x86), then reinstalling from scratch. Same result. Have also tried both the included private JRE and installing the latest JRE manually from Sun.... nothing seems to change this startup problem.

Thanks,
DK

[BillP · Accepted solution]

ok..got a few more questions.

you start the service - this seems as though its reproducible all the time? after a reboot, manually, starting/stopping, correct?

also, when you start the PCNS service and then check the PCNS event log when you are able to access the software - is the time stamp for the entry PowerChute Network Shutdown version 3.0.0 started indicating it started when you actually start the service (before the 10-15 min access delay) or is the time stamp for that event indeed 10-15 minutes later when the service becomes available?