server up!

[Portasaurus]

Originally Posted by Rogean [You must be logged in to view images. Log in or Register.]

Heap Corruption with a popup that prevents the batch file from restarting the process because it doesn't fully crash out (main stream keeps running). Kanras and I have spent days worth of time working on going through code from the time it started happening (September - November 2011) checking and rolling back code and haven't found the cause.

Frankly at this point it is the single biggest thing pissing me off, more so than even the latest drama troll scandal.

http://www.rogean.com/images/worldcrash.png

When this happens it just starts spamming connection fails because the UDP Thread of the process is no longer executing, which is also the reason it stops responding to any UDP packets from clients for logging in or zoning.

Some folks have tossed out band-aid solution ideas and I might as well add to the list while I wait:

I've had an issue where a program would "hard-crash" and pull up th e"this program has stopped responding" window, which is actually some form of the "Doctor Watson" crash handling program.

When that happened, I needed to clear the "this program has stopped responding" window automatically, so I just set up a scheduled task to run "KILL DWWIN.EXE" EVERY minute (or whatever the process was that opened when I got this unwanted pop-up). This way it would completely CLOSE out of the crashed program once the debugger popped up, which would then allow the other scheduled task that tries to run the batch file every minute (as long as that task wasn't already currently running, an option in task scheduler).

Net result is: whenever a crash happens, it closes out of the program automatically, and then the automated scheduled task re-opens my batch file that launches the program.

[Hollywood]

Originally Posted by Portasaurus [You must be logged in to view images. Log in or Register.]

Some folks have tossed out band-aid solution ideas and I might as well add to the list while I wait:

I've had an issue where a program would "hard-crash" and pull up th e"this program has stopped responding" window, which is actually some form of the "Doctor Watson" crash handling program.

When that happened, I needed to clear the "this program has stopped responding" window automatically, so I just set up a scheduled task to run "KILL DWWIN.EXE" (or whatever the process was that opened when I got this unwanted pop-up). This way it would completely CLOSE out of the crashed program, which would then restart from a 2nd scheduled task that was just continually trying to run that program over and over (as long as the task wasn't already currently running).

Net result is: whenever a crash happens, it closes out of the program automatically, and then the automated scheduled task re-opens my batch file that launches the program.

One would hope it's something that simple but this is a three step process - ideally

A) Bring on board some European guides to reset server when we are asleep
B) Use a batch like you've suggested to reset server if it's crashed/down for more than thirty minutes at a time(e.g. noone was able to manually reset)
C) Fix the problem itself

In that order should be their priority until it's sorted.
BUT RE: the European thing, sometimes I wonder if that would really make a difference as it's noon now and yet it's still not fixed. Surely not all GMs with reset access are still asleep, already gone out or died last night.