Enough is Enough

[Sabot1]

Well, tonight I officially got Pizzed at the game, and logged out.

Not because of the game play or the banter, but because of the CTD/CTHL or whatever you want to call it.

Crashing because you were killed in the middle of an exchange is unacceptable.

The game won't respond, you restart the game, and everything is completely locked, so you have to restart the game again.

Not acceptable..........Rats, you can do better......

[ltdata]

I have not had that type of problems in a very long time. Maybe you should give your system a check? Clean up the harddrive?

[dasspeil]

After a crash / CTD, restart the game then after its just connected (while loading) alt F4 then reload.

for some reason, after a crash / CTD, the game needs 2 restarts to work properly.

[kepta0]

I have been having the same problem lately. Crash then two restarts to work.

[khbigjoe]

Everyone has this problem.

Very rarely do you log back in straight away with no problems.

Either do the alt f4 thing or wait a miute or two before logging back in.

I totally agree though..shocking that this is a common problem

[alfheim]

I posted about this in an effort to warn CRS they would lose half of their new subscribers within the first week if this was their experience; a moderator locked the thread.

Well, the problem hasn't gone away.

(Interesting my thread about the same problem was locked, and not yours.)

Hope they get this sorted out before 1.31 goes live.

[Topd]

Well, tonight I officially got Pizzed at the game, and logged out.

Not because of the game play or the banter, but because of the CTD/CTHL or whatever you want to call it.

Crashing because you were killed in the middle of an exchange is unacceptable.

The game won't respond, you restart the game, and everything is completely locked, so you have to restart the game again.

Not acceptable..........Rats, you can do better......

Did you check and see if your CTD resulted in an error log in the BE folder?

[werwolf]

This seems to affect some people heavily, others once in a while, and a few not at all. Very strange.

[MOTORMOUTH]

MOving this to where you can actually get some help.

[Scking]

I rarely get a CTD in 1.30 if at all.. I had a squadmate that constantly CTD'ed as well and someone told him to reboot his router and he hasn't CTD'ed since.. Don't always blame the game.. There is alot of equipment between you and CRS's machines

[lure]

My personal opinion is that a big portion of the CTHL problems with this game is related to (one of?) the ISP CRS/Playnet uses for the game servers, CogentCo. I did extensive testing a few months back while I was almost completely cut of from the game due to the very same ISP blocking traffic from my ISP.

CocentCo is a crap ISP and they have years of history of getting into trouble with other ISP, using traffic blocking to/from other ISPs as some kind of negotiation technique. Several times before have CRS customers been completely cut of from the game servers due to CogentCo having some fight with some other ISP and starting to block traffic.

It also seem that CogentCo have very overbooked routers. I often see lots of packetloss occuring at CogentCo routers.

CRS/Playnet need to seriously consider moving to a different ISP but I bet that can be expensive and maybe there are contracts involved. Who knows.

As I'm typing this I have been running a "My traceroute" log against the IP 66.28.6.226 that is labeled "Primary network" in the wwiiol_netcheck.bat script that comes with BE (located in game folder) and over a period of 15 minutes it shows a 31% packet loss at the router "te4-1.ccr01.phl01.atlas.cogentco.com" located very close to the "Primary network" IP above. 31% packet loss is totally unacceptable and will ensure a very poor game experience for any customer forced to play through that route.

Sure, some CTHLs are the result of poor ISPs at the player end but I'm sure that the main part of the CTHL complaints and reports are caused by poor CogentCo service quality.

[Sabot1]

Did you check and see if your CTD resulted in an error log in the BE folder?

Sorry about the rant Rats,

Its was just frustrating to let my squad down at a very critical moment.

Concerning a CTD, I can almost anticipate when they will occur. It usually happens right after a "Hard" kill. What I mean is something like sitting on a grenade, or getting shot and you seem to go through a building, or eating 10 bofors rounds as you dive on a FB. Last night I was inf and was hit and went down into a tree. This was the third time I was taken out in as many minutes. Don't know if there is any correlation here.

Now concerning a log, I checked the folder and there was no log generated. But I have a question concerning logs. How long do they stay in your folder? If you log right back in, will it erase the log? The only reason I am asking is because I did generate a log in the Beta, and I checked the folder, and that log is now gone.

And Motor, you are probably right.....I did need some help last night....mental help....

Edited March 5, 2010 by Sabot1

[lutorm]

My personal opinion is that a big portion of the CTHL problems with this game is related to (one of?) the ISP CRS/Playnet uses for the game servers, CogentCo. I did extensive testing a few months back while I was almost completely cut of from the game due to the very same ISP blocking traffic from my ISP.

I thought they had two different ISPs?

[lure]

I thought they had two different ISPs?

The reason to have two ISPs should be to have one available as backup "just in case" and not to have a extra ISP because the main ISP suck.

For 3-4 month I had NO contact with the primary server. The cutoff point was inside CogentCos network. CogentCo actively blocked traffic coming from the network providers my ISPs used. This was the 2nd time I was cut off from CRS/PLaynet by CogentCo.

Solution? I went through 3 different ISPs to find one that was not blocked by CogentCo. I now have a choice to use primary or secondary depending on where the connection is better. By looking at the traces and traceroute logs during my testing I'm very sure that a big number of CRS/Playnet customers suffers from the poor quality of the CogentCo network.

[lure]

Oh, it looks like I've kept the log running for the last hour+ now. I have only 10 jumps from my ISP here in northern parts of Sweden to the CRS "Primary network" including my own router at home. You have to consider that pretty good. Pings and Loss are good all the way in Europe and over the sea and into CogentCos network. There we enter packet loss galore.

Traceroute with pings sent every 10 seconds to each router along the way:

[DOC]

Nice Lure, I'll see that this gets passed along to Killer. It's a nightmare to be sure, but thanks for giving us your findings. I cannot comment on what will happen next.

[KFS1]

Oh, it looks like I've kept the log running for the last hour+ now. I have only 10 jumps from my ISP here in northern parts of Sweden to the CRS "Primary network" including my own router at home. You have to consider that pretty good. Pings and Loss are good all the way in Europe and over the sea and into CogentCos network. There we enter packet loss galore.

Traceroute with pings sent every 10 seconds to each router along the way:

Traceroute works by sending out a 'ping' ("ICMP_ECHO_REQUEST") packet with a limited "time to live" (TTL); each network devices that sees this request lowers the time to live by 1. When it reaches zero, that device sends you back a Host Unreachable message.

So by starting with a TTL of 1, you get Host Unreachable from the very first network device on the way to the destination.

However, network devices are under no obligation whatsoever to honor an ICMP_ECHO_REQUEST, or even to pass them on.

For high-end network switches and routers, processing an ICMP_ECHO_REQUEST can be incredibly expensive.

Traceroute is thus only a guide. It's like testing how deep a hole is by throwing sticks of dynamite. ("Well, it *was* 9ft deep"

It's also worth noting that your traceroute first shows packet loss at "gblx.net". Which could either mean:

a. A bad edge-router at rksnet.se,

b. A peering issue between riksnet.se and gblx.net,

c. A bad edge-router at gblx.net,

d. A busy edge-router at gblx.net that ignored some of your ICMP_ECHO_REQUESTS,

That means that any results beyond the gblx.net hop are invalidated. You can't actually tell what's going on past a hop with packet loss. Even if you are using a traceroute that directly pings individual hosts.

Then there is 0.9% packet loss on the first hop into cogent. This is very likely to be the same packet-loss source as the gblx.net issue and seems likely to indicate a high-load router at gblx.net.

I suspect that your traceroute application monitors individual hops on the route, by pinging them after using the traceroute method for finding hosts along a route. And the hop 10 listed on your traceroute is one of a series of load-balancing routers. So if you are trying to ping it directly, yeah: you're likely to "see" packet loss with ping/icmp packets.

I don't currently see any cogent issues from here or from my servers in the UK.

But it's worth noting that we have two ISPs: If you are having connectivity issues via cogent, use the game's settings tool to switch to the secondary network.

- Oliver

[lure]

Yes the linux tool mtr (My Traceroute) works that way: It's a combination of traceroute and ping. Of course ICMP echo is considered a low priority but if such packets get dropped even when sent with long intervals in between it's a good indication of something bad going on. Some routers start dropping packets by design if you hammer them with echo requests packets too often and that's why I'm testing with 10, 20 or even 30 seconds intervals.

Of course the 2-3% along the way have to be considered when looking at the 30% loss but it can't nearly explain it all. A 2-3% drop of ICMP echo packets don't feel like a big problem when it comes to playing BE. But IMO if a router start dropping 1/3 of the echo requests it's gotta be under a good deal of load and it is much more likely that other packets are dropped as well.

Also, I can tell you guys one more thing: The packetloss in the cogentco network shows in the netcode2 packetloss indicator also. Low packetloss in cogentco on the ping log = low packetloss in game, high packetloss in ping log = high packetloss in game.

The BIG issue however is that CogentCo is manually blocking traffic from certain ISPs. The packets (ALL kinds of packets) reach the first CogentCo routers but get dropped with the reply "Destination network unreachable". As the destination network is CogentCo that is kindof clear that they are dropping the traffic. I doubt a router would be left with broken config for month after month.

This result that you have customers with ONLY option to play on the second link no matter how bad it is.

I was in that situation for several month and was forced to play over a route that took my packets on a sight-seeing of half of Europe. This is from the ISP i was using in second half of last year (while NO connection to "Primary network" at all):

Packets Pings

Host Loss% Snt Last Avg Best Wrst StDev

1. 192.168.1.1 0.0% 939 0.3 0.4 0.3 21.3 1.4

2. 213.21.110.1 0.0% 939 4.8 2.4 1.3 22.2 2.9

3. 88-83-32-253.core.t3.se 0.0% 939 1.6 1.9 1.3 31.9 2.3

4. 88-83-32-254.core.t3.se 0.0% 939 2.5 3.0 1.2 18.4 3.2

5. svl-s-10ge.core.t3.se 0.0% 939 4.9 5.3 4.8 25.1 1.6

6. sthlm-s-10ge.core.t3.se 0.0% 939 11.0 13.9 10.9 30.1 3.7

7. static-213-115-153-249.sme.bredbandsbolaget.se 0.2% 939 11.3 13.1 11.2 201.9 11.4

8. ti3001d400-xe1-0-0.ti.telenor.net 0.3% 939 22.2 15.7 11.4 78.3 11.5

9. ti3001c310-ae1-0.ti.telenor.net 0.4% 939 21.7 26.3 21.7 80.7 8.7

10. ti3004c310-ae6-0.ti.telenor.net 0.2% 939 22.1 26.3 21.7 85.1 8.5

11. ti3004b300-ae0-0.ti.telenor.net 0.9% 939 22.4 25.0 21.6 64.5 6.6

12. 212.73.252.1 1.4% 939 23.7 26.3 22.3 110.3 11.0

13. ae-11-11.ebr1.Berlin1.Level3.net 0.7% 939 36.2 39.4 33.7 55.9 4.7

14. ae-1-100.ebr2.Berlin1.Level3.net 0.6% 939 37.9 39.7 33.8 55.3 4.6

15. ae-3-3.ebr2.Dusseldorf1.Level3.net 0.5% 939 40.6 41.5 40.3 55.3 2.1

16. ae-2-2.ebr1.Amsterdam1.Level3.net 0.0% 938 43.8 45.1 43.2 59.0 2.7

17. ae-1-51.edge4.Amsterdam1.Level3.net 6.8% 938 44.3 48.0 43.5 142.2 12.0

18. 208.174.49.69 0.7% 938 44.8 45.9 43.8 109.1 6.2

19. dcr1-so-1-0-0.amsterdamamx.savvis.net 0.9% 938 44.7 50.4 43.8 190.3 17.5

20. cr2-pos-0-0-3-0.londonuk1.savvis.net 0.3% 938 59.3 59.7 58.4 111.2 2.7

The above route was unplayable by any standard I'm used to. The packetloss with following warping and CTHL made the game experience so poor that I almost didn't play at all for 3 months or so.

What I'm trying to say is that the main problem with CogentCO is that they are blocking god know how many ISP from their networks and they keep doing that for MONTHs. So players who probably would have the best route to the server through CogentCo are forced to use a secondary route that might be very poor.

You should not use a ISP that keep blocking customers from your servers.

[GOPHUR]

There is a ticket for this #3092.

I'll elevate the status a bit to see if I can get it fixed. It's a bad exit opint for a player.

My more immediate concern is why are you crashing all the time. I play 6 hours a day and I never crash. We need to investigate this.

How often do you crash? Where do you crash?