Mini-Lags

[tar]

Since several days (weeks?) there randomly occur regular mini-lags. Every 5-10 seconds (or even worse) there is a 0.5-1 second freeze or even "connection interrupted" that interrupts the movement, gameplay, aim, everything. This occurs not in every game, but mostly around 19-23 o'clock. Is there an issue with the server?

[fernandinho1337]

Have you checked the lagometer? Do you see red spikes?
My routing was broken for a long time and still is. If I traceroute the server in the mornings it is perfect (just 10ms more ping than usual) but if I play in the evenings I have packetloss on a regular basis.

Best regards

*edit* As I understood adminless, it is an issue Deutsche Telekom (DTAG) <--> OVH. Is DTAG your provider?

[tar]

1) Of course I see red and yellow spikes. It is definitively a server/connection issue.
2) If you experience the exact same issues I had yesterday evening, how were you able to get 93 points in the same match? I did not observe signs of any lags on your side.
3) I use a local cable provider since 2009 and never had such problems before.

[fernandinho1337]

1) well OK then your connection is screwed. I guess adminless can see more details. maybe message him directly
2) do you really want an explanation?
3) I use the offline version of HLSW. you can enter your favorite servers and do a tracerouting. it then graphically shows you if you have packetloss and where packetloss starts. for example I can tell that my connection is fine until the data stream reaches the "intersection" from DTAG (my provider) to OVH (server provider). so thats why I can tell where problem could be (and of course, adminless told me, too).

best regards

[adminless]

hey hello there, yes although you're purchasing your connection form a third party, your provider is just a DTAG re-seller (or in other words, your provider is basically running on the DTAG network). that being said, there's absolutely nothing wrong with DTAG as a company or with its infrastructure, DTAG overall professionalism and quality is totally out of question here, DTAG has been know to work consistently without any major issues at least since I've been running top servers like this (this is, more than a decade), it just happened that as with anything from time to time there's always some issues here and there and yes as fer is telling you this time happened that DTAG has been suffering from low quality routing at its internal network at least for the last month, so it's likely that that's what is affecting you know too now. it has been much worse though, I remember that when fer first complained it was totally unplayable, up to 90% packet loss, although apparently after a few complains they shorted it to some extend at least. probably your best option there is to casually complain (basically just report) about low quality connection to them as that's a bigger issue than just a simple end user connection issue.

as for the server itself as I see it now is performing as usual, basically flawlessly, the maximum frame delay time in the last month didn't even reach 150 msecs at max and that's even when changing maps and the rest of the time frames get transmitted within mere msecs with just some tens of them delayed bellow 100 msecs over millions of them a day (i.e. +99.998% of real availability). and as for ovh itself there's also no know issues at the moment either that I'm aware of, as I see it is running as usual (i.e. quite good too). all the problem seems to be just that, DTAG, as you can see in this graph here. that link is ovh reverse diagnosis as you can see there if you look up other networks (example) DTAG is probably not the only one of course but of the few that is dirty at the moment.

sorry for the inconvenience, hope things get sorted soon.

[tar]

@fernandinho

2) I assume, you did not by experience any connection issues within that particular game.
3) I just downloaded it and I attached the trace "log". But I don't know what to do with that. It swings from 30 to 140 ms.

@adminless

That is not only annoying - it is not playable anymore. I'll get in contact with my provider.

Thanks, both.

[fernandinho1337]

@tar

interestingly at some point your connection goes completely red.
what i find really weird is that your ping is already >50ms at your fritzbox. is there anything wrong with your connection overall?

hit WINDOWS-KEY + R, enter "cmd" and do "ping heise.de" and "tracert heise.de" - what do you see there? maybe ur connection is screwed overall

2) i have less packetloss thank you do but on a regular basis. when my ping goes completely down the drain i just stop playing. i scored 93 yesterday because i rely more on railgun and play less aggressive since i have those issues. my connection issues actually made me a better player/ made me go back to old playing style. i guess you are very well aware that playing railgun is a lot more efficient than shafting

best regards

[tar]

I contacted my provider and they switched me to a private IP - but I do not think that solved anything as HLSW still goes to red. I also did ping & tracert two times, see attachments.

Is my connection screwed? Do you think my FritzBox is the reason?

[fernandinho1337]

when you do the tracerouting to heise.de your ping in the end is good. what i would like to know

if you do "ping heise.de -n 100" does any of the packets get lost? if not it is really a routing issue to the unfreeze server.

i also found my routing/ping to unfreeze2 is much better (is it a different provider). can you check how that one is? just out of interest

best regards

[tar]

Most between 20 and 40 ms. One on 60 and one on 159 ms, none loss:

Code: Select all

Ping-Statistik für 193.99.144.80:
Pakete: Gesendet = 100, Empfangen = 100, Verloren = 0 (0% Verlust),
Ca. Zeitangaben in Millisek.: Minimum = 20ms, Maximum = 159ms, Mittelwert = 26ms

For server.unfreeze.ga its even better:

Code: Select all

Ping-Statistik für 217.182.45.102:
Pakete: Gesendet = 100, Empfangen = 100, Verloren = 0 (0% Verlust),
Ca. Zeitangaben in Millisek.: Minimum = 21ms, Maximum = 103ms, Mittelwert = 26ms

And for server2.unfreeze.ga:

Code: Select all

Ping-Statistik für 51.255.170.73:
Pakete: Gesendet = 100, Empfangen = 100, Verloren = 0 (0% Verlust),
Ca. Zeitangaben in Millisek.: Minimum = 22ms, Maximum = 220ms, Mittelwert = 29ms

Its always only one outlier.

[tar]

OK, I just connected and played a bit.... and it was smoother than never before: no packetloss and constant ping. Therefore, I will test it again this evening and if I have issues again, I will do another ping + tracert.

[tar]

i guess you are very well aware that playing railgun is a lot more efficient than shafting

I do not agree here as shafting is the most spamming thing one could do and is especially high effective for low ping players. Railgun is more efficient on high ranges - even when I use to rail on close combat

[fernandinho1337]

@newb,.zi here you can compare my connection

best regards

[adminless]

mmmm they switching your ip makes sense, in fact, as I see it now, yes everything is solved now for you as they actually changed you from the (at the moment) DTAG dirty network to a ("unknow" to me, first time I heard of) "envia TEL GmbH" network (looks like they're re-sellers themselves or have partnership with various top providers not just DTAG) so you're no longer going trough the DTAG network, hence you're problems are gone now (which clearly confirms that the part at blame was indeed DTAG and not the server side). to be honest that's some hell of a epic response from them, props for them for solve your issue that fast, so that good, happy it got solved then

[tar]

Ok, now the lag is back... lets do some pings:

Code: Select all

Ping-Statistik für 217.182.45.102:
    Pakete: Gesendet = 100, Empfangen = 100, Verloren = 0
    (0% Verlust),
Ca. Zeitangaben in Millisek.:
    Minimum = 20ms, Maximum = 942ms, Mittelwert = 45ms

No packetloss, but some > 500 ms.

Code: Select all

Ping-Statistik für 217.182.45.102:
    Pakete: Gesendet = 100, Empfangen = 100, Verloren = 0
    (0% Verlust),
Ca. Zeitangaben in Millisek.:
    Minimum = 21ms, Maximum = 2090ms, Mittelwert = 59ms

There is it... 2 seconds! Now tell me, you can hit shit with that lag. Lets try some trace:

Code: Select all

Routenverfolgung zu server.unfreeze.ga [217.182.45.102]
über maximal 30 Hops:

  1    <1 ms    <1 ms    <1 ms  fritz.box [192.168.178.1]
  2     *        *        *     Zeitüberschreitung der Anforderung.
  3    10 ms    17 ms    14 ms  192.168.0.30
  4    82 ms    14 ms    12 ms  ip-212-99-206-65.enviatel.net [212.99.206.65]
  5    11 ms    17 ms    17 ms  be3-rb2-c.envia-tel.net [77.235.191.225]
  6     *        *        *     Zeitüberschreitung der Anforderung.
  7    26 ms    26 ms    23 ms  be101.sbg-g2-nc5.fr.eu [91.121.215.196]
  8     *        *        *     Zeitüberschreitung der Anforderung.
  9    24 ms    22 ms    20 ms  be7.sbg-z2g1-a75.fr.eu [37.187.232.53]
 10     *        *        *     Zeitüberschreitung der Anforderung.
 11    22 ms    22 ms    19 ms  51.255.191.108
 12    25 ms    27 ms    21 ms  164.132.232.5
 13    22 ms    29 ms    21 ms  server.unfreeze.ga [217.182.45.102]

Ablaufverfolgung beendet.

Looks good, but:

trace3.png