By your own figures you are experiencing the leak, just because your
not aware of it doesn't mean it doesn't exist.
Please don't confuse the situation by spouting uninformed superstition.
Regards
Steve
- Original Message -
From: "Ryan" <[EMAIL PROTECTED]>
> This must be a
oh that gave me a good laugh
--- On Sat, 6/28/08, Steve Stifler <[EMAIL PROTECTED]> wrote:
From: Steve Stifler <[EMAIL PROTECTED]>
Subject: Re: [hlds] TF2/OB memory leak
To: "Half-Life dedicated Win32 server mailing list"
Date: Saturday, June 28, 2008, 1:15 PM
Valve doesn't post on this list e
This must be a server specific thing though, we have 3 TF2 servers on
one Windows 2003 box, and we have never experienced this. Our servers
have all been up for days completely full of players 32/32. Our 2fort
server runs for 600 minutes between map reloads, Goldrush and Dustbowl
90 minutes and 300
Valve doesn't post on this list except to say "Bend over, here somes another
update!!!"
On 6/28/08, Benjamin Lange <[EMAIL PROTECTED]> wrote:
>
> There is no leak.
> A memory leak leads to an increasing amount of memory without new events
> happening.
> As we can see, this not the case.
> srcds is
There is no leak.
A memory leak leads to an increasing amount of memory without new events
happening.
As we can see, this not the case.
srcds is simply not giving resources free, after once taken them.
Servers near 1GB of memory usage are definitely not acceptable after just
2-3days.
Best solution
As for the memory leak on map change. Most of my servers run the
same map and never once change map. maxrounds 50.. I see the
leak progress as people leave/join the server. As the player ID
goes up so does the ram on my box. On day 3 its nothing to see
the server sucking up 1.5G and not change m
Memory leaks every map change.
This is not effected by the number of players that join / leave the server,
only happens at map change.
My graphs top have the max memory set to 500 megs, and here is an image of the
graphs showing 2.5x that amount.
So far this has not caused any crashes or performa
Andrew Armstrong wrote:
> We definitely see TF2 servers leaking memory.
>
> Our server graphs show that when the server first starts (around 6am) memory
> usage remains constant.
>
> As soon as players join the server, memory begins to rise as they remain on
> the server.
>
> When the server become
To be clear, my graphs (2f2f) were both from Linux TF2 servers.
Ryan
On Sat, Jun 28, 2008 at 3:30 AM, Saint K. <[EMAIL PROTECTED]> wrote:
> And here is the mem leak on Linux;
>
> http://www.specialattack.net/downloads/uptime_15h.JPG
> http://www.specialattack.net/downloads/uptime_35h.JPG
>
> Sain
And here is the mem leak on Linux;
http://www.specialattack.net/downloads/uptime_15h.JPG
http://www.specialattack.net/downloads/uptime_35h.JPG
Saint K.
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Ryan Mannion
Sent: Saturday, June 28, 2008 11:14 AM
T
Yep thats about it I'm sick of servers that suck up 1G or more
Please fix this Valve.
___
To unsubscribe, edit your list preferences, or view the list archives, please
visit:
http://list.valvesoftware.com/mailman/listinfo/hlds
This is a big problem, and it's only existed for the past week or so.
Take a look at http://www.2fort2furious.com/images/2f2f_memusage.png -
the trend is very obvious. Note the time frame - only five hours.
Though changing maps is obviously correlated with a significant spike,
even a 24/7 server
We definitely see TF2 servers leaking memory.
Our server graphs show that when the server first starts (around 6am) memory
usage remains constant.
As soon as players join the server, memory begins to rise as they remain on
the server.
When the server becomes empty, memory remains constant at whe
13 matches
Mail list logo
