Tideswell Mass DC during NW

mynamewasstolen · September 2015

This is getting ridiculous. Lately a lot of players have been having extremely high ping rates, who never experienced such before the servers were merged. Ping rates are anywhere between 0 and 22k for myself. When I have 0 ping all I can do is chat and nothing else. When I have 22k ping im just sitting there looking at everyone stand in place for a long time. Now, the issue has carried over into nation wars and caused what I'm only guessing is about half the player population to be disconnected from the game and upon relogging getting a message "unable to enter max capacity instance" - effectively getting booted from the NW upon being relogged. Why is there not any way to get a second NW entry pass when stuff like this happens? Admittedly that would be a great way to alleviate the obvious server issues we are having.

sylenthunder · September 2015

Did you take a traceroute while you were having this issue?

alaia#4592 · September 2015

._.

mynamewasstolen · September 2015

I know that's your typical response to server dc issues Sylen, but I think it screams server issue when you log back in and there is at least 40 other people spawning back in at wellspring village also complaining about being disconnected and booted from NW. It isn't a localized ISP issue either. I'm on Time Warner in NC, and I know one of the other people who got booted is on Comcast in PA, and others were connecting from overseas.

sensualsoul · September 2015

Distinguished Players:
The last exit location of your character >name here< in five Emperors. Unable to enter Max Instance reashed...

Srsly PWI, when if ever are you going to get your **** pile working right? I think you need to give us another NW to do to make up for us missing this one. NOT only am I out the tokens I should of made but I'm out the apo, NW apo and ink dragon apo and such that I use only for NW.

< I can't speak for the other but I'm really tired of this!

sylenthunder · September 2015

mynamewasstolen wrote: »

I know that's your typical response to server dc issues Sylen, but I think it screams server issue when you log back in and there is at least 40 other people spawning back in at wellspring village also complaining about being disconnected and booted from NW. It isn't a localized ISP issue either. I'm on Time Warner in NC, and I know one of the other people who got booted is on Comcast in PA, and others were connecting from overseas.

It screams regional hub issue, but until you guys start posting data, we'll never be able to tell. If all you ever do is cry about the issue, we'll never narrow down where it actually is.
If we cannot narrow down the source definitively, how do you ever expect anything to get fixed?

mynamewasstolen · September 2015

if traceroutes help, then im sure a traceroute posted from now would help just to satisfy the curiosity whether 40 people all went through the same hop that might've had an issue rather than PWI itself. However, it seems to me the issue lies with the server since these high ping rates didn't start for me and others until the merge happened. It seems as though the server can't handle the higher traffic load it is experiencing now after the merge, and ends up dropping connections during peak play times. If I knew the new url/ip for the tideswell server I could post one definitively. as for now, pwieast2.perfectworld.com will have to work:

[code]

Tracing route to pwieast2.perfectworld.com [198.49.243.16]
over a maximum of 30 hops:

1 10 ms 9 ms 9 ms 192.168.0.1
2 32 ms 26 ms 28 ms cpe-65-184-160-1.ec.res.rr.com [65.184.160.1]
3 44 ms 41 ms 39 ms 174.111.108.33
4 14 ms 41 ms 28 ms cpe-024-025-055-234.ec.res.rr.com [24.25.55.234]
5 51 ms 31 ms 54 ms be35.drhmncev01r.southeast.rr.com [24.93.64.192]
6 50 ms 44 ms 39 ms bu-ether15.asbnva1611w-bcr00.tbone.rr.com [66.109.6.80]
7 36 ms 44 ms 49 ms 0.ae2.pr0.dca10.tbone.rr.com [66.109.6.169]
8 48 ms 39 ms 43 ms be3021.ccr41.iad02.atlas.cogentco.com [154.54.10.209]
9 52 ms 46 ms 49 ms be2171.ccr41.dca01.atlas.cogentco.com [154.54.31.105]
10 47 ms 51 ms 47 ms be2148.ccr41.jfk02.atlas.cogentco.com [154.54.31.118]
11 62 ms 72 ms 54 ms be2094.ccr21.bos01.atlas.cogentco.com [154.54.30.14]
12 68 ms 50 ms 58 ms te0-0-1-0.rcr11.b002133-1.bos01.atlas.cogentco.com [154.54.46.134]
13 57 ms 49 ms 49 ms 38.111.40.114
14 64 ms 59 ms 59 ms 198.49.243.253
15 * 75 ms 47 ms 10.251.192.21
16 * * * Request timed out.
17 64 ms 61 ms 59 ms 198.49.243.16

Trace complete.

[/code]

jadasia · September 2015

Jump 16 timed out.

If it was a server issue everyone would have ended up at Wellspring Village after logging back in because of the map crash. The fact that a fraction of those who DC from PWI it does seem like a node issue instead of a server issue.

mynamewasstolen · September 2015

and heres for pwieast1.perfectworld.com

Tracing route to pwieast1.perfectworld.com [198.49.243.15]
over a maximum of 30 hops:

1 11 ms 10 ms 9 ms 192.168.0.1
2 28 ms 32 ms 31 ms cpe-65-184-160-1.ec.res.rr.com [65.184.160.1]
3 44 ms 49 ms 55 ms 174.111.108.33
4 32 ms 28 ms 31 ms cpe-024-025-055-234.ec.res.rr.com [24.25.55.234]
5 36 ms 39 ms 49 ms be35.drhmncev01r.southeast.rr.com [24.93.64.192]
6 65 ms 63 ms 40 ms bu-ether15.asbnva1611w-bcr00.tbone.rr.com [66.109.6.80]
7 35 ms 58 ms 33 ms 0.ae0.pr0.dca10.tbone.rr.com [66.109.6.183]
8 58 ms 40 ms 39 ms be3021.ccr41.iad02.atlas.cogentco.com [154.54.10.209]
9 58 ms 49 ms 40 ms be2171.ccr41.dca01.atlas.cogentco.com [154.54.31.105]
10 48 ms 57 ms 64 ms be2148.ccr41.jfk02.atlas.cogentco.com [154.54.31.118]
11 56 ms 51 ms 61 ms be2094.ccr21.bos01.atlas.cogentco.com [154.54.30.14]
12 69 ms 61 ms 87 ms te0-0-1-0.rcr11.b002133-1.bos01.atlas.cogentco.com [154.54.46.134]
13 57 ms 64 ms 59 ms 38.111.40.114
14 56 ms 65 ms 69 ms 198.49.243.253
15 57 ms 59 ms 48 ms 10.251.192.21
16 * * * Request timed out.
17 53 ms 52 ms 50 ms 198.49.243.15

Trace complete.

the last hops have always timed out though since their own routers block icmp packets

nrglg · September 2015

I've been testing and data collecting. One of PWE's East ISP's LAN nodes is having problems. As far as I can tell, only east coast and European players will have issues accessing the east coast server Tideswell. West coast people won't have any problems funny enough (I didn't do any central USA testing). Hopefully PWE's ISP will resolve their issues soon enough. This would explain why a great majority of people are being kicked. However, to reword what I'm saying, the server is fine, but an ISP node that most, if not all of you are going though, is not.

sylenthunder · September 2015

@mynamewasstolen Your issue begins before you even get out of your house. Your first hop is from your PC to your modem, and that should never be more than 3ms, 5ms if you're using Wifi. Hitting 10-11ms right off the bat shows that you have an issue between your PC and your modem. Either is's a weak wifi signal, a poorly insulated cable, lots of interference (wired or wireless), or a software/hardware issue.

jadasia wrote: »

Jump 16 timed out.

If it was a server issue everyone would have ended up at Wellspring Village after logging back in because of the map crash. The fact that a fraction of those who DC from PWI it does seem like a node issue instead of a server issue.

Because you don't see a huge increase in the ms response after that hop, it just means that the device at hop 16 is not returning ping requests.
If it was indicative of an issue at that particular hop, typically the numbers after it would be much larger.

You are correct though in that if it was a server issue, it would affect all, not a few. If it were just the NW map, it would be everyone in the instance, which also wasn't the case.

mynamewasstolen · September 2015

sylenthunder said:
@mynamewasstolen Your issue begins before you even get out of your house. Your first hop is from your PC to your modem, and that should never be more than 3ms, 5ms if you're using Wifi. Hitting 10-11ms right off the bat shows that you have an issue between your PC and your modem. Either is's a weak wifi signal, a poorly insulated cable, lots of interference (wired or wireless), or a software/hardware issue.

so youre saying that about 40 other people just coincidentally happened to have the same issue as me at the same time as me? I'm not really buying that sylen. maybe I do have a somewhat weak connection to my wifi router, but I don't drop off my network and im also able to stream HD video on Netflix without it stopping to buffer all the time. PWI takes a lot less data transfer than that. Theres something more going on than just "my issue" as you put it.

sylenthunder · September 2015

mynamewasstolen wrote: »

@mynamewasstolen Your issue begins before you even get out of your house. Your first hop is from your PC to your modem, and that should never be more than 3ms, 5ms if you're using Wifi. Hitting 10-11ms right off the bat shows that you have an issue between your PC and your modem. Either is's a weak wifi signal, a poorly insulated cable, lots of interference (wired or wireless), or a software/hardware issue.

so youre saying that about 40 other people just coincidentally happened to have the same issue as me at the same time as me? I'm not really buying that sylen. maybe I do have a somewhat weak connection to my wifi router, but I don't drop off my network and im also able to stream HD video on Netflix without it stopping to buffer all the time. PWI takes a lot less data transfer than that. Theres something more going on than just "my issue" as you put it.

If you read my second post, and what NRGLG wrote, it appears to be a internet node near the server.

What I mentioned won't cause you to DC often, but it will cause you to DC. You don't have a bunch of pre-buffered data playing an MMO like you do watching a video online. (Just because Netflix doesn't stop to buffer, doesn't mean that there weren't lost data packets, just that they weren't enough to overcome the buffer cache.)
I didn't say that it was the only part of your issue, but it is a part of your issue you can do something about. It would also be a major cause for your ping to experience heavy spikes in the game.

Again I will state this, as I have numerous times over the past 6 years. The cause of the "mass DC" is most likely due to a hiccup on a large internet gateway.

ishtera1970 · September 2015

Hello Sylenthunder,

I would like to point out that when I had problems last week I placed the trace route in a thread and was never commented to. That being said. My husband and I game together with him sitting less than two feet from me. We have been doing this for TWO years without ever having this problem. Your point of it being a hub issues is funny to me considering he was one of the ones in the mass dc yet I was able to complete NW. So if it is a hub issue, why did I not dc? We share the same modem, we are both hardwired to it, his computer is newer and faster than mine, and yet he dced. As everyone keeps pointing out, these high pings and dcs did not begin occuring to the majority of us unti AFTER this merger. As for the trace route, I am not sending another one as you ignored my last one and considering that he and I are both online right now with NO problems a trace route 3 hours after the fact will do you no good. Have a nice day.

sensualsoul · September 2015

@sylenthunder It sounds like you want the players to do the GMs job for them. You say "cry about the issue" I say it's not MY job to figure out where the problem is, it's the GMs they get paid to keep this game run smoothly. It is, us the players that buy zen that pay them for a service and they suck at doing their jobs. The East Coast server has had issue for YEARS, and we did get told that the merge they did would fix it, it did not. So the GMs lied once again. So can you blame us for "crying" about it? If they bothered to monitor the game they might see where the problem is! I don't give a flying damn how it works, I want to spend my time in the game without the dc's and lag, not in forums posting to begin with! Let alone doing someone elses job.

sylenthunder · September 2015

ishtera1970 wrote: »

Hello Sylenthunder,

I would like to point out that when I had problems last week I placed the trace route in a thread and was never commented to. That being said. My husband and I game together with him sitting less than two feet from me. We have been doing this for TWO years without ever having this problem. Your point of it being a hub issues is funny to me considering he was one of the ones in the mass dc yet I was able to complete NW. So if it is a hub issue, why did I not dc? We share the same modem, we are both hardwired to it, his computer is newer and faster than mine, and yet he dced. As everyone keeps pointing out, these high pings and dcs did not begin occuring to the majority of us unti AFTER this merger. As for the trace route, I am not sending another one as you ignored my last one and considering that he and I are both online right now with NO problems a trace route 3 hours after the fact will do you no good. Have a nice day.

It's very common for the packets for one data stream to get dropped while another will not. Which explains exactly how one PC can get disconnected while the other doesn't.

We've been saying for years that you need to take the traceroute at the time you are having the issue. maybe the issue is with load balancing on a router near the server, or even with PWI's ISP.
Unfortunately we will never be able to find out until we start getting hard data instead of people complaining and not giving any solid information.

People have been complaining about latency issues for forever. They have just been louder about it since the merger.
They're also not providing information.

@sensualsoul the GM's can't run traces to your home PC's to find the source of the latency and dropped signal. Only you can do that. Monitoring the game is just going to show that you guys are disconnecting due to a break in the data stream. If enough of you guys provide the data we've been asking you to provide for years, then maybe an issue can be proved. Until then it's just a bunch of people complaining that the internet is behaving normally.

nrglg · September 2015

ishtera1970 wrote: »

I would like to point out that when I had problems last week I placed the trace route in a thread and was never commented to.

I commented on your trace and another person's simultaneously. Go back and see the thread. I was pointing out similarities in the path both of you were taking and that you both had the same ISP. This meant, from my point of view, that there was a strong possibility that you were both using the same DNS server which could have been giving you all lousy paths.

sylenthunder · September 2015

You know, NRGLG and I have both done a lot, and gone out of our way, to help you guys understand the reality of how the internet works, and to give you the tools you need to help us help you.

Please please please, is it at all possible for everyone posting to just stop the whine and start working to give us valid information to work with? I mean seriously, we see hundreds of these threads, and only a small handful of people give us anything that we can use to help you, or to narrow down a common cause.

mynamewasstolen · September 2015

sylenthunder said:

Please please please, is it at all possible for everyone posting to just stop the whine and start working to give us valid information to work with?

Now that I am done playing, let me please refer you back to my very first tracert > tracert.txt output I copied and pasted for you here:

Tracing route to pwieast2.perfectworld.com [198.49.243.16]
over a maximum of 30 hops:

1 10 ms 9 ms 9 ms 192.168.0.1
2 32 ms 26 ms 28 ms cpe-65-184-160-1.ec.res.rr.com [65.184.160.1]
3 44 ms 41 ms 39 ms 174.111.108.33
4 14 ms 41 ms 28 ms cpe-024-025-055-234.ec.res.rr.com [24.25.55.234]
5 51 ms 31 ms 54 ms be35.drhmncev01r.southeast.rr.com [24.93.64.192]
6 50 ms 44 ms 39 ms bu-ether15.asbnva1611w-bcr00.tbone.rr.com [66.109.6.80]
7 36 ms 44 ms 49 ms 0.ae2.pr0.dca10.tbone.rr.com [66.109.6.169]
8 48 ms 39 ms 43 ms be3021.ccr41.iad02.atlas.cogentco.com [154.54.10.209]
9 52 ms 46 ms 49 ms be2171.ccr41.dca01.atlas.cogentco.com [154.54.31.105]
10 47 ms 51 ms 47 ms be2148.ccr41.jfk02.atlas.cogentco.com [154.54.31.118]
11 62 ms 72 ms 54 ms be2094.ccr21.bos01.atlas.cogentco.com [154.54.30.14]
12 68 ms 50 ms 58 ms te0-0-1-0.rcr11.b002133-1.bos01.atlas.cogentco.com [154.54.46.134]
13 57 ms 49 ms 49 ms 38.111.40.114
14 64 ms 59 ms 59 ms 198.49.243.253
15 * 75 ms 47 ms 10.251.192.21
16 * * * Request timed out.
17 64 ms 61 ms 59 ms 198.49.243.16

as you can see here plainly, once you move past my 10, 9, 9 MS pings to my wireless router - the 15th hop to the game server (10.251.192.21) dropped the first ping request on my tracert command. The asterisk denotes such per your own example tracert posted in your tech support thread here: http://forum.arcgames.com/pwi/discussion/comment/6592708/#Comment_6592709 under "Tracert to Heavens Tear wrote:" - at the 9th hop. This IP address is a private class A IP address, located beyond the 14th hop router of 198.49.243.253. Doing a "who is 198.49.243.253" search on the internet brings up the following: https://www.findip-address.com/198.49.243.253 a further search on "Cryptic Studios", to whom this IP is registered to - brings up the following: https://en.wikipedia.org/wiki/Cryptic_Studios. The evidence is quite clear in all of this that the 15th hop on the route between my PC and the game server is dropping packets, and further more is located within PWE's own routing system. I've done my research, and I hope this helps. Furthermore, I think I speak for the community at large when I say that instead of constantly insisting the issue lies with the user - try to apply some logic and look at all the evidence presented. This was an issue experienced by not just myself, not just coincidentally 2 or 3 others - but by a multitude of people at once. We weren't all in the same geographic region in order to blame it on the hub our ISP uses, much less the individual ISP. Again I state that these extreme lag issues did NOT occur prior to the merge - yet now they do, and nothing has changed at the home since then, and I'm sure the same holds true for many other players. On top of this, I have been dc'd mid NW before and was able to log back in without being booted from NW - yet this time 40 or so people experienced the EXACT same issue. I don't believe this is a client end error or hub error if that is occurring. Please try to account for all the evidence presented, and I'll try to provide you with a tracert to Tideswell the next time the extreme lag spikes occur - if you can give me the current proper IP or URL for that server. I can also provide the same by activating the hotspot on my droid instead, with the phone much closer to my wifi antenna than my router is.

sensualsoul · September 2015

@sylenthunder I have sent a number tickets over the years with info such as tracing routes that was ignored or as of the Arc being used I been told to post in forums. Why the ticket system no longer can be use for making a ticket in reguards to a game issue is beyond me but I NOT like to post in forum since people like you feel the need to "help me understand" that even though it IS a PWE issue you feel it is something with my computer or internet or whatever. I do not want to deal with you or any number of trolls on forums, nor did I ask for your help in those threads, I was posting there for the GMs! I stopped posting anything since no GM ever replied just people like you to whom I was not making the post to. So why should I waste my time? If the GMs would look at the tickets or even a thread once in a while and REPLY to the players more of them might give them what THEY ask us for. YOU are not a GM however and can NOT do a damn thing about our issues so we are wasting our time even talking to you. So now like many, many players I use the forums to let of my ire instead since yelling at my screen is not as satisfying as telling the GMs how much they suck at their jobs!

zoner112012 · September 2015

Everyone bashing sylenthunder when he is the only person on forums that actually offers help to players when he really doesn't have to need to wake up. When you tick him off enough he will just stop offering to bust his balls to help you. sylen knows wth he is talking about. I would trust his word over a GM in most circumstances. Also, those saying he is asking you to do the GMs job for them aren't using your head. A GMs job is in game issues, and has nothing to do with your connections to their servers. Again, stop being pricks and listen to sylen and maybe he can help you, but if the issue IS related to a node between you and the server all you can do is hope the ISP finds the issue and fixes it or try to report it for them. Either way, everyone's anger and words toward sylen being all ticked off with what he is telling you isn't helping you and in the end could end up hurting us all. He has always been the most helpful person on forums, you don't bite the hand that feeds you.

nrglg · September 2015

@sensualsoul I've known people who worked as GMs at different MMO companies (I don't think I can say which). My knowledge of networking surpassed theirs (which was pretty much nonexistent) and I'm not a networking expert and consider it to be my weakest topic of knowledge in computing. So asking GMs to reply to your networking questions is pointless. Most of them have no freaking idea how to read and understand your traces. And like sylen has said so many times before, a large portion of the time it is not PWE. This is just a fact. It's not a defense or PWE or anything like that. It is cold hard facts based on actual evidence (or a lack thereof, because so many of you are resistant to do anything but whine).

@mynamewasstolen You're jumping way, way, ahead of yourself. You jumped to conclusions based on lack of understanding (and evidence for that matter). Most of what you said needs to be answered individually because your conclusions were just plain wrong and need to be smacked out of your head. Especially about understanding why you could suddenly be having problems with, "nothing has changed at the home since". Guess what, cables and splitters are not magic and have a finite lifespan. The sun can also interfere believe it or not. So never, ever, jump to this conclusion again. That is just ignorance at an unacceptable level. For mild home maintenance you should know this. However, going back to the primary single trace provided. I would recommend instead of doing a single trace after the fact and such and then coming to a final conclusion based on a single trace, that you monitor the game server while you're playing each day and save the output. That would give not only a much clearer picture of where the issue appears to be, but could definitively prove that one of PWE's routing devices is at fault if you're correct. If you want to do that, get a program called PingPlotter. I would give you my program, but frankly I'm still working on it at the moment. So PingPlotter is one of the best choices for that at the moment. Getting long-term data about this would give you the evidence needed to criticize and complain with validity.

TL;DR, read it. I gave you a way to silence anyone who doubts you if you are in fact correct. So prove PWE is at fault.

sylenthunder · September 2015

@sensualsoul NRGLG pretty much summed it up. I'd also like to add that as much as I say that it's usually not PWE, and as much as it looks like I'm defending them... If I had hard data that showed it was, I would bring my full resources to bear down on them. (not that it would likely do much good these days, but it's better than nothing.) The truth of the matter though, is that 98% of the time the issue is completely not PWE, and when we're given hard data, we're able to determine that.
I am not a GM for PWE, although I have GM'ed before. That doesn't mean that I cannot do anything about it though. I still have contacts at PWE and Wanmei.

Also, when we're given hard (and accurate) data, we're typically able to find the source of the issue and help you find a resolution. If you think that's a waste of time...
In the end, it works out a hell of a lot better than if you just sit around and whine your **** off. (no offense intended)

kittyempressa · September 2015

Sylen, this wasn't just your typical "mass DC during NW". In every NW-mass DC before yesterday's the dc'd char was in nation base when logged back in, like it should. Also, in the case of connection-based mass dc's there's usually a short time before server shows online to the client again. But...

This time people got dc'd(note: only people in NW dc'd, Kitty didn't hear of anyone on main map dc'ing) midfight and got the issue mentioned in first post of arcgames.com/en/forums/pwi/#/discussion/1202976/nw-bug . The characters were loggable instantly after dc unlike in the usual cases.

Depiction of the problem: At some point the in-game stoofs stopped responding, akin to your usual "server error"-issue. Soon "You've been disconnected"-message appeared. When logging back in postimg.org/image/moz08rey7/ (image from that other thread) appeared. If you pressed "Confirm" character logged in, at Wellspring Village instead of nation base, still with nation marker above it's head and nation chat visible(like all those "Light nation has occupied E8"). Kitty also heard someone say that trying to re-enter at Aeban only gave "Leave Nation War"-option so server probably thought people were still in NW though they were on main map. And for comparison: if you use teleport stone to get outside NW, you retain nation marker but you can't see nation chat.

That happened to (according to Kitty's estimation based on people appearing in Wellspring about every 3 secs for 5 mins) about 80-100 characters, maybe more.
Kitty thinks it must be heavily related to that "max instances reached" which would point towards instance server having issues as client used to say that when server couldn't open more instances during those old "kill FB19 boss"-events.

sylenthunder · September 2015

kittyempressa wrote: »

Sylen, this wasn't just your typical "mass DC during NW". In every NW-mass DC before yesterday's the dc'd char was in nation base when logged back in, like it should. Also, in the case of connection-based mass dc's there's usually a short time before server shows online to the client again. But...

This time people got dc'd(note: only people in NW dc'd, Kitty didn't hear of anyone on main map dc'ing) midfight and got the issue mentioned in first post of arcgames.com/en/forums/pwi/#/discussion/1202976/nw-bug . The characters were loggable instantly after dc unlike in the usual cases.

Depiction of the problem: At some point the in-game stoofs stopped responding, akin to your usual "server error"-issue. Soon "You've been disconnected"-message appeared. When logging back in postimg.org/image/moz08rey7/ (image from that other thread) appeared. If you pressed "Confirm" character logged in, at Wellspring Village instead of nation base, still with nation marker above it's head and nation chat visible(like all those "Light nation has occupied E8"). Kitty also heard someone say that trying to re-enter at Aeban only gave "Leave Nation War"-option so server probably thought people were still in NW though they were on main map. And for comparison: if you use teleport stone to get outside NW, you retain nation marker but you can't see nation chat.

That happened to (according to Kitty's estimation based on people appearing in Wellspring about every 3 secs for 5 mins) about 80-100 characters, maybe more.
Kitty thinks it must be heavily related to that "max instances reached" which would point towards instance server having issues as client used to say that when server couldn't open more instances during those old "kill FB19 boss"-events.

That may be entirely possible that it is the server instance overloading, but typically when that happens, it will dump everyone out, not just a few of them.
It's most likely that the network path to that particular server isn't managing the load balancing very well when there is the heavy load in that instance with all of the people joining it. Unfortunately, in order to determine whether that is on PWE's end, or the internet, we need data.

The issue with the nation title and chat is a bug in the code, and should be noted in the Quality Corner forum.

kittyempressa · September 2015

sylenthunder wrote: »

That may be entirely possible that it is the server instance overloading, but typically when that happens, it will dump everyone out, not just a few of them.
It's most likely that the network path to that particular server isn't managing the load balancing very well when there is the heavy load in that instance with all of the people joining it. Unfortunately, in order to determine whether that is on PWE's end, or the internet, we need data.

The issue with the nation title and chat is a bug in the code, and should be noted in the Quality Corner forum.

Considering that we've had overloading issues earlier (port/whatever on PWE's end allowing some to connect, but not everyone. PWE admitted it happened at their end that time), could it also apply in this case so that instance server has thrown out people until it has been able to handle the remaining connections?

It's just that Kitty has hard time believing the issue would be with the interwebz as all the tracert's posted in this thread have been normal. If the issue was with the interwebz, nodes before 4th hop before the server(last node) should show ping issues. Not to mention that everytime it was interwebz issue, Kitty hasn't been able to log back in right away (and those times Kitty knew it was interwebz issue from tracert showing ping issues before Cryptic Studios. And in yesterday's "Tideswell taaaun"-case Kitty couldn't get ping from server, but from every node before that incl. Cryptic's first 2 nodes)).

shimarra · September 2015

Welcome to DW server guys! It's always had more crashes than other servers... Has anyone suggested to check your firewall yet?

sylenthunder · September 2015

kittyempressa wrote: »

sylenthunder wrote: »

That may be entirely possible that it is the server instance overloading, but typically when that happens, it will dump everyone out, not just a few of them.
It's most likely that the network path to that particular server isn't managing the load balancing very well when there is the heavy load in that instance with all of the people joining it. Unfortunately, in order to determine whether that is on PWE's end, or the internet, we need data.

The issue with the nation title and chat is a bug in the code, and should be noted in the Quality Corner forum.

Considering that we've had overloading issues earlier (port/whatever on PWE's end allowing some to connect, but not everyone. PWE admitted it happened at their end that time), could it also apply in this case so that instance server has thrown out people until it has been able to handle the remaining connections?

It's just that Kitty has hard time believing the issue would be with the interwebz as all the tracert's posted in this thread have been normal. If the issue was with the interwebz, nodes before 4th hop before the server(last node) should show ping issues. Not to mention that everytime it was interwebz issue, Kitty hasn't been able to log back in right away (and those times Kitty knew it was interwebz issue from tracert showing ping issues before Cryptic Studios. And in yesterday's "Tideswell taaaun"-case Kitty couldn't get ping from server, but from every node before that incl. Cryptic's first 2 nodes)).

I'm pretty sure I already stated in this thread, but will do so again just in case.....

You will almost never get a ping response from the server. Because DDoS. Same goes for PWI's direct ISP link.

kittyempressa · September 2015

sylenthunder wrote: »

You will almost never get a ping response from the server. Because DDoS. Same goes for PWI's direct ISP link.

According to tracert that's not true. pwieast2.perfectworld.com's IP-address returns ping, while the hop before it (presumably login server) doesn't. Unless Tideswell-server's address isn't really located in pwieast2.perfectworld.com (which would quite illogical since datafiles show 29000:pwieast2.perfectworld.com as Tideswell's server address). If you don't believe, check those tracerts above.

From mynamewasstolen's post:
"Tracing route to pwieast2.perfectworld.com [198.49.243.16]"
"17 64 ms 61 ms 59 ms 198.49.243.16"

Thus, as long as 198.49.243.16 returns ping during tracert, server's supposedly online and connectable. In that case there shouldn't be too many possible causes for spontaneous dc/inability to log on, other than packages missing, server made unloggable(happens usually during maintenances) or some problem in PWE's systems(like port/gateway/whatever going bad).

The reason Kitty assumes the hop before 198.49.243.16(server) to be login server would be that if one wanted to DDoS, easiest way to make server unavailable would be attacking login server to prevent logging in, Kitty thinks (though she's just an amateur in this stuff).

thesquee1 · September 2015

sylenthunder said:

It screams regional hub issue, but until you guys start posting data, we'll never be able to tell. If all you ever do is cry about the issue, we'll never narrow down where it actually is.

If we cannot narrow down the source definitively, how do you ever expect anything to get fixed?

It was definitely a hub issue. Funny enough when I logged and decided to play MC the server there had the same issues. That's a dead giveaway, too, when it's not just PWI.

jadasia · September 2015

Here's mine to HH server:

Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.

C:\Users\Acer>tracert pwigc2.perfectworld.com

Tracing route to pwigc2.perfectworld.com [66.151.133.71]
over a maximum of 30 hops:

1 5 ms <1 ms 1 ms REPOSITORY [192.168.1.1]
2 1 ms 1 ms 1 ms 192.168.0.1
3 23 ms 33 ms 19 ms boid-dsl-gw15.boid.qwest.net [184.99.64.15]
4 21 ms 19 ms 19 ms boid-agw1.inet.qwest.net [184.99.65.113]
5 43 ms 43 ms 51 ms snj-edge-04.inet.qwest.net [67.14.34.82]
6 43 ms 44 ms 43 ms 63.145.225.174
7 44 ms 43 ms 44 ms border2.t7-1-bbnet1.sje004.pnap.net [66.151.144.
20]
8 44 ms 44 ms 58 ms perfectworld-10.border2.sje004.pnap.net [64.95.1
43.190]
9 45 ms 43 ms 54 ms 66.151.133.71

Trace complete.

Not sure why it went to my USB flash drive that is plugged into my router but do notice just as Sylen said jump 2 is from my PC to my router via WiFi at 1 ms. You really do appear to have some major bottleneck between your computer and the router. It's nearly 10 times slower than it should be.

Tideswell Mass DC during NW

Comments

Categories