Server issues: New server

[Xcom]

Have anyone noticed any pattern to the server performance?

Today at around 10 pm est the server started to become much worse then its usual laggy self. But the player count was at roughly 80. Lower then the usual 130. Anyone noticed any pattern to the lag? player count online, specific timezones, days? Trying to at least stay clear of these timezones to not get a headache. I can tolerate some lag but right now the server is lagging twice as much as its unfrozen.

Just a note for Loftar if he wants a good sample for testing. At the time of posting this the lag is at its peek and its been lagging like this for about an hour.

[borka]

There where sooo many guesses on patterns already ...

but we won't be able to get a real clue about it anyways as we all would have to be able to post "our personal time" correctly in UTC /GMT (some ppl even don't know whether it's EST or EDT... ) to be measurable, also we would have to have knowledge about how networks work and how to determine the individual factors (do you know how many times a day your provider is switching routes your data is going?) ...

i.e.

EST 10PM is GMT 2.00am
EDT 10PM is GMT 3.00am

Germany is at UTC/GMT +2 atm

so EDT 10PM is 5AM next day in germany

now an example of "my personal network factors":

time 11:45 UTC/GMT - connection via UMTS / 3G
tracert moltke.seatribe.se

1 my PC to Provider network

2 608 ms 398 ms 398 ms 82.113.122.198
3 356 ms 398 ms 398 ms IARMUN1-Gi0-2-199.net.de.o2.com [82.113.122.2]
4 366 ms 398 ms 398 ms xmwc-mnch-de01-gigaet-5-1-510.nw.mediaways.net [195.71.164.209]

5 356 ms 398 ms 418 ms decix-gw.hetzner.de [80.81.192.164]
6 347 ms 418 ms 398 ms hos-bb1.juniper2.rz12.hetzner.de [213.239.240.252]
7 356 ms 398 ms 398 ms hos-tr4.ex3k15.rz12.hetzner.de [213.239.228.240]
8 376 ms 418 ms 418 ms moltke.seatribe.se [178.63.100.209]

2 - 4 my Providers network
4 - 5 Transfer point (my providers network - Hetzner network)
5 - 8 Hetzner network - rz12 is where the server is located, hos are switches

now imagine this ping + servers lag (client uploading but not receiving data) = stop motion HnH but i won't see a difference when player count changes

I know certain reasons for my high ping like the provider has a jamming in my area atm...etc.

[factnfiction101]

I'm having a 45 second lag with 4-5 second little lags, It's playable for 6 seconds at a time. Currently it's 8:35 (GMT -5:00) EST The lag started at 6am, I think

It isn't playable...

[Xcom]

At times like these can't one of the moderators on the forums reboot the servers? Its sad if only Mr bossman can do that. Would be nice if the mods had the power to push the reboot server button for rare instances like this one.

Edit: wow the lag just cleared up out of the blue

I swear some guy at Hetzner uses the server as a bouncing wall plank. Its the only explanation that fits.

[Cranny]

When it is like this I wish it to crash so we get some relief.

[Shask]

Fix game pl0x. It doesn't matter if you've got to reset the server, start a new world, wipe everyone's character, etc.. just stop the lag I BEG YOU!!

Although as of the exact time of writing, the lag seems to have gotten a lot better..

[zacty]

Although as of the exact time of writing, the lag seems to have gotten a lot better..

fuckin told you guys

[Oddity]

Fix game pl0x. It doesn't matter if you've got to reset the server, start a new world, wipe everyone's character, etc.. just stop the lag I BEG YOU!!

Although as of the exact time of writing, the lag seems to have gotten a lot better..

Code: Select all

(15:56:38) <loftar> Well, my basic problem in the issue is to find out just what is happening. I mean, I know that the cause of the lag is in basic I/O calls, but what actually happens is all in the kernel.
(15:56:48) <loftar> So I'd need to figure out some way to see what's happening in the kernel.
(15:56:57) <loftar> Just what resource the thread in question is blocked for.
(15:57:32) <loftar> I've considered trying out LatencyTOP, but it requires kernel patching and stuff and I still just don't really have the time to do that.
(15:57:33) <loftar> QQ

QQ indeed

[mvgulik]

As far as I'm concerned there seems to be a general pattern. And that is that after a server restart the lag is significant less for some time. (I'm not talking about the initial period where the player count is still building up to its general level.)
After the initial return of the general lag, things seems to degenerate slowly over time.

But considering this is currently a abandoned project. There is no real point to suggest a automated daily server restart to see if this might be true. (Just in case ... NOTE: this is not the same as a automated server restart after a server crash.) (+ I doubt there is some saved data available on the dev's side to do a more objective compare.)

Personally
... I think that they should mothball this project ...

[zacty]

As far as I'm concerned there seems to be a general pattern. And that is that after a server restart the lag is significant less for some time. (I'm not talking about the initial period where the player count is still building up to its general level.)
After the initial return of the general lag, things seems to degenerate slowly over time.

But considering this is currently a abandoned project. There is no real point to suggest a automated daily server restart to see if this might be true. (Just in case ... NOTE: this is not the same as a automated server restart after a server crash.) (+ I doubt there is some saved data available on the dev's side to do a more objective compare.)

Personally
... I think that they should mothball this project ...

i think the devs will be returning here soon, since salem is starting to get closer to open beta