On Mon, Jul 30, 2012 at 1:36 AM, Russell Smith ve...@tinylittlerobots.uswrote:
My turn to think out loud.
In the more extreme cases of this, players will get disconnected with
STEAMAUTH: Error code 6. Does anyone know what this error code actually
means? If the issue were between the
k_EDenyInvalid = 0,
k_EDenyInvalidVersion = 1,
k_EDenyGeneric = 2,
k_EDenyNotLoggedOn = 3,
k_EDenyNoLicense = 4,
k_EDenyCheater = 5,
k_EDenyLoggedInElseWhere = 6,
k_EDenyUnknownText = 7,
k_EDenyIncompatibleAnticheat = 8,
k_EDenyMemoryCorruption = 9,
k_EDenyIncompatibleSoftware = 10,
Hmm, that's odd that a logged in elsewhere error would be thrown. Is
there a unique identifier assigned to each server when handshaking with
Steam? Perhaps if the server lost connection to Steam long enough for
this identifier to be reset when a new handshake occurs it would look
like all
Hmm, that's odd that a logged in elsewhere error would be thrown.
It was the wrong enum, error 6 means the client cancelled the ticket,
i.e. they quit the game or left the server.
~
Their heads are green, and their hands are blue,
And they went to sea in a Sieve. - Edward Lear
On
: [hlds_linux] Random CPU Spikes
Hmm, that's odd that a logged in elsewhere error would be thrown.
It was the wrong enum, error 6 means the client cancelled the ticket,
i.e. they quit the game or left the server.
~
Their heads are green, and their hands are blue,
And they went to sea
The entity data seems normal for a payload map. The client
extrapolation seems odd as my understanding is that this would only
occur when the client fails to receive updates from the server. But the
net graph seems to indicate the server is still sending out tic updates
since there are no
I don't think it's entities that cause that, especially not in Badwater.
Most entity heavy map that i know made by Valve is Thundermountain and
it isn't even near the engine limits, even with the dynamic ones created
on the fly.
I'm pretty sure the issue is hardware related. It would be
What would you like to know about the CPU usage? As the thread title
implies srcds is spiking to 100% utilization of a core. Load averages
in htop increase accordingly.
I find it hard to believe that multiple servers running on different
OSes entirely would suddenly all start having
Just thinking out loud. I ment the CPU usage over the whole machine at
that time when it happens. You could try to use strace on the process
that is chewing up the core, see what it is doing.
I don't think it's hardware either if it's multiple machines involved
are having the issue but i've
Besides the srcds thread that is spiking, the rest of the processes seem
to behave normally taking up a reasonable amount of CPU time. I have a
Left 4 Dead 2 server running on the same box which is having no issues
as far as I can tell.
What's the performance penalty for using strace? I
My turn to think out loud.
In the more extreme cases of this, players will get disconnected with
STEAMAUTH: Error code 6. Does anyone know what this error code actually
means? If the issue were between the player and server then I would
expect to see a timeout message instead of a steamauth
These are still a thing. They started around the time of the pyro
update. The main srcds thread spikes to 100% cpu utilization and all
players in game will lock in place until CPU usage goes back to normal.
Users will see something like this on the net graph in the more extreme
cases:
Is it completely random or during someone joining?
On Sat, Jul 28, 2012 at 12:21 AM, Russell Smith
ve...@tinylittlerobots.uswrote:
These are still a thing. They started around the time of the pyro update.
The main srcds thread spikes to 100% cpu utilization and all players in
game will lock
It doesn't happen on player joins. I haven't been able to see any
commonalities between occurrences that would indicate what the cause is.
On 7/28/2012 12:31 AM, doc wrote:
Is it completely random or during someone joining?
On Sat, Jul 28, 2012 at 12:21 AM, Russell Smith
I see this too. I think it has something to do with entities because of the red
spike in net graph but I'm not sure.
Sent from my iPhone 4
On Jul 28, 2012, at 12:33 AM, Russell Smith ve...@tinylittlerobots.us wrote:
It doesn't happen on player joins. I haven't been able to see any
On 28.07.2012 08:21, Russell Smith wrote:
normal. Users will see something like this on the net graph in the
more extreme cases:
I haven't had any users running the netgraph command, but I am getting
complaints in chat about lagspikes from time to time, and it has been
more prevalent since
If the server is assigned to one core only with taskset, you will see spikes. I
used to have them like that but valve made changes while back. Now especially
on map changes, server can easily chew a lot more cpu than normally. You also
need to have atlest 600 megs memory free in order to have
On 28.07.2012 10:29, ics wrote:
If the server is assigned to one core only with taskset, you will see
spikes. I used to have them like that but valve made changes while
back. Now especially on map changes, server can easily chew a lot
more
But, CPU spikes on map change would not give spikes
Map change/round change would give spike, if you run multiple
gameservers in same machine and there isn't enough free cpu to provide
juice. It wouldn't affect the gameserver where the cpu goes up
temporarely but some of the other gameservers, it might.
Last time i had similiar issues, it was
I'm not locking srcds instances to a single core. I did try locking
each to two cores as a solution but that had no effect, so I removed
that restriction.
I'm also not running using -threads 1. I did try that as there was a
thread on this list two years back that mentioned that as a
20 matches
Mail list logo
