Forums » General
I don't want to start a new post, because I have no idea how long the server will be down for. However, this is the second down time.....
It looks like this is the time the new server is gonna be installed but no one is sure.
Well, it's back up. We're using the new hardware and the new code. The new code is continuing to blow up in new and interesting ways, so I'm afraid things will be a little unstable like they were in the 3.3.0 time period until this code fully matures.
Sorry about that, but hopefully our growing pains will be over with next week.
Sorry about that, but hopefully our growing pains will be over with next week.
We upgraded from a single PIII 550mhz with 384mb of ram running FreeBSD 4.3 to a dual Athlon MP 2400+ with 1gb of ram running Linux 2.6.1. OS switch to take advantage of Oprofile (kernel profiler available in Linux). New hardware seems to be cool and stable, we've been testing it here at the office for some weeks. The new server code is still exploding periodically, and that's what we'll have to sort out over the next few days.
It's spazzing out again.....
It crashed just as I destroyed a vulture with swarm missiles.
It crashed just as I destroyed a vulture with swarm missiles.
Hmm, i figured you'd be in that sector camps... But yah, it's out again.
SIgh, gone again, and this time it means i probably issued:D
Yeah, I just fixed that problem. That particular bug seems to be in Linux, so we never saw it under FreeBSD. I receive a file descriptor from select() as readable, then when I recvfrom() on it, it gives me EAGAIN. So.. it's lying to me, or something. Anyway, that case is handled, but there are a couple baffling crashes to fix (which are the same as what I was seeing on the previous server).
If I put another CPU in my machine, mine would be better than yours. :P
Yeah, well, it was the most powerful machine we could buy for the least amount of money.
Been booted again. :(
http://www.cs.helsinki.fi/linux/linux-kernel/2002-29/0271.html
/me yells at the server and tells it it will be if it doesn't let us play it will be grounded!
hmm, if its running inside valgrind that'd explain why things timeout before you can log in... *G* valrgind isn't the.. .erm.. fastest.. of production environments.
Good luck to the devs in debugging this, personally I hope they work over this weekend so I can blast some bots, but I won't hold it against them if they take the weekend off.. (okay, maybe I will, but just a little)
[edit]
And thanks for the update on whats running here, incarnate
Good luck to the devs in debugging this, personally I hope they work over this weekend so I can blast some bots, but I won't hold it against them if they take the weekend off.. (okay, maybe I will, but just a little)
[edit]
And thanks for the update on whats running here, incarnate
Ugh, back up again... I tried to fix some obscure network code bug last night that valgrind found, but I fixed it improperly, so the server wouldn't accept any connections when it finally segfaulted and restarted with the new code. Back up now. Still haven't fixed what seems to be the actual problem, but hopefully this afternoon I will.
andreas: That would be the bug. Not a big deal really.
andreas: That would be the bug. Not a big deal really.
Server died again.
And returned again.
Good choice with the Dual Athlon MPs, the AMD 64-bit stuff has dropped the price of the MPs nicely, and they offer a very good performance/price ratio... thinking of getting one of these systems myself now... ;-)
Could we have our stats saved more frequently? I racked up alot of score yesterday in 17, and didn't jump.. when the server went down i lost it :( Perhaps every 10 mins or something?
Could we have our stats saved more frequently? I racked up alot of score yesterday in 17, and didn't jump.. when the server went down i lost it :( Perhaps every 10 mins or something?
I think the server just went down again.
and back up.
and back up.
Ehehe, the last kick that i experienced warped me back to before AKA had killed me once:P