Weird Problem... (Page 1/1)
Cliff Pennock APR 04, 06:02 PM
For some strange reason, if you tried to get in to the "Technical Discussion", your browser would eventually time-out without displaying the section's contents. I had closed the forum for a while, trying to figure out what was wrong but the problem solved itself after a while. I still have no clue what happened so please let me know when it happens again. Sorry for any inconvenience.
Monkeyman APR 04, 09:53 PM
No problems now!
DRH APR 05, 12:09 PM
Hmmm... This may not be a new thing. I had that problem about a month or so ago. It was late after midnight CDT. I tried several times and could not get to the Technical Forum although I could still get to the rest. Everything was fine the next day so I didn't worry too much about it.
theogre APR 05, 02:12 PM
well, If this was a normal database doing that, I'd look at what else is happening on the system(s) at the time. Especially if it's repeatable.

That forum is the biggest one. That might have something to do with it. The software might be bogging down handling it. If that's the case then maybe we need to look at archiving some of the older threads, especially big threads like "clutch info uncovered." Archives would make that stuff allot easier to find. ((I've noticed we are seeing a real increase in the number of repeated questions that have already been heavily covered.))

Might be related to search. That's been really slow at times. Could be hogging the hardware. One thing that might help there is altering the search time ranges. Changing them to 5, 30, 90 days, maybe 120 days, from the current 5, 10, 30. That would help shorten the searches. There are allot of searches where you want to go back more than 30 days, but not search since the begining of time.

Also I'd look at what time does the backup job runs and how the job runs when it does. That can really screw up server performance.

Cliff Pennock APR 05, 03:23 PM
I don't think that section is too big, simply because when it does work, it's pretty fast and it doesn't seem to have any troubles at all. I think it has something to do with the program trying to fork (yes, that is an existing term!) another process that has reached its maximum instances... Anyway, I'm still trying to figure this one out.
theogre APR 06, 02:19 PM
Myself I can only handle 2 forks. Maybe a fork and a knife. Spoons give me the creaps but you can't eat soup to well without them......

I don't pretend to know this server well..... Doesn't linux have a thread monitoing tool? Preferably one that you could set alarms in? Then you'd know real quick if that's what's happening. It would seem someones got a linux one like that. ((Now Ogre remembers why he stopped being a programmer.......))

I haven't had time to fool with linux. Could it be a swap memory problem? I know SQL on NT has traditionally thrashed the hell out of the swap when it starts using more than a few pages. Linux is likely better but it could be something similar if Ultimate or the CGI processor codes got a bug.