[bf1942] Latest build observations

[g8 at the.whole.net]

I am also seeing the unusually high pings.  Everyone that joins my server
has a very high ping, usually well above 100, and it tends to fluctuate
quite a bit.  This has been tested with people that would normally have
sub 100, relatively stable pings to the same server/network (and it's at a
datacenter so no problems there).  Like a few others, I also still get
connection problem notices when it gets above 10 or so players.  My pub
test serevr seems to attract a lot of players so i've been able to observe
performance with quite few people connected.

And for those that are interested, here the ps output from a fully loaded
32 player server running on a dual 2ghz P4 with FreeBSD 4.6-Stable:

11:55am|root at u1[~]#
USER         PID %CPU %MEM   VSZ  RSS  TT  STAT STARTED      TIME COMMAND
nobody     40574 73.0  5.9 138588 124516  p0  R+    7:52PM 179:25.14
./bf1942_lnxded +game BF1942 +restart 1 +dedicated 1 +reconnectPassword 6719
+reconnectSlots 26 +config ServerSettings.con

-g8

On Tue, 7 Jan 2003, Anton Jansen wrote:

> Hi all,
>
> I would like to share with you my experiences with the latest build of
> the last 24 hours. Seems that the crashing of the server is gone, I have
> not had a single core dump, with the previous build it was something about
> 2-3 a day. Played Wake Island for about 2 hours yesterday, the connection
> problem window seems to have dissappeared, now only the pings of the
> players rise.
>
> On closer observation I can see that when I'm alone on the server I got a
> ping of 58 instead of 34, which I had with the previous build. (A "normal"
> ping has a time of 32ms). In the previous version my ping was not only
> lower, but also more stable (less deviation), now with the latest build I
> got maximum pings of 158. It has nothing todo with CPU load (which is
> always lower than 0.7) or network bandwith. The only thing I could thought
> up was the excessive debugging generating additional i/o, which makes the
> overall system slower, but I'm not sure ...
>
> Can people confirm my observations?
>
> --
> Best regards,
>
> Anton Jansen