Lag / network failures, traced to ZOS

fred4

Dear ZOS, over the past weeks I have experienced periodic lag spikes. This has not happened as persistently over the previous 2 years. They can last several minutes and occur maybe once every 2 hours. Sometimes they occur more frequently. During the spikes, the game feels unresponsive and jerky. At worst - when the game didn't outright disconnect - combat was frozen for 5 seconds at a time. I believe this may be a problem with ZOS systems and would be grateful if you could relay the following information to your network specialists.

I looked up the IP-addresses the game connects to using my personal firewall. I then performed a traceroute, as follows:

1 1 ms 1 ms 1 ms 192.168.1.254
2 31 ms 31 ms 30 ms b-ras2.prp.dublin.eircom.net [159.134.155.6]
3 32 ms 39 ms 31 ms tenge-9-1-2.pe1.prp.bbh-prp.eircom.net [86.43.247.169]
4 33 ms 30 ms 30 ms lag-102-coreb-prp-pe1-prp.coreb.prp.core.eircom.net [86.43.9.168]
5 47 ms 46 ms 46 ms lag-20.br1.thw.border.eircom.net [86.43.13.240]
6 45 ms * 44 ms ldn-b1-link.telia.net [62.115.50.13]
7 46 ms 46 ms 46 ms ldn-bb2-link.telia.net [62.115.143.26]
8 43 ms 44 ms 43 ms ldn-b4-link.telia.net [62.115.134.139]
9 47 ms 47 ms 47 ms dtag-ic-323085-ldn-b4.c.telia.net [62.115.50.19]
10 65 ms 61 ms 60 ms 217.239.45.238
11 61 ms 60 ms 60 ms 87.190.232.70
12 60 ms 61 ms 60 ms 195.122.154.3
13 * 61 ms 61 ms 159.100.232.134

I then performed continuous pings to the game server address 159.100.232.134. When I experienced in-game lag spikes, I observed intermittent ping timeouts to that IP-address. I also performed pings to the following addresses at the same time:

217.239.45.238: There were no ping timeouts to this address, which belongs to Deutsche Telekom AG.
87.190.232.70: This address, belonging to Deutsche Telecom AG does not respond to pings, e.g. it is probably configured to ignore them.
195.122.154.3: There were ping timeouts to this address, which according to public records may belong to Zenimax.

In conclusion, it appears there are periodic problems with some of your network infrastructure or servers, specifically your connection to Deutsche Telekom AG. It is also possible there is a problem within Deutsche Telekom. Since 87.190.232.70 never responds, it is unfortunately not possible to determine whether the problem lies there. All the same I would be grateful, if you could investigate. Thank you.

fred4

This problem is going on right now, and has been for at least 15 minutes.

fred4

I had a prolonged lack of network communication 10 minutes ago (14:30), putting my character back quite a distance, in Cyrodiil. I was subsequently just about to reach Imperial City, when the game threw me to a loading screen that looks like the one you get before character selection, e.g. not a Cyrodiil loading screen. It's been sitting there for a few minutes now. Pings, meanwhile, are back to normal with no timeouts for a few minutes. Today's IP-trace is:

1 1 ms 1 ms 1 ms 192.168.1.254
2 33 ms 32 ms 30 ms b-ras2.prp.dublin.eircom.net [159.134.155.6]
3 32 ms 32 ms 34 ms tenge-9-1-2.pe1.prp.bbh-prp.eircom.net [86.43.247.169]
4 33 ms 31 ms 31 ms lag-102-coreb-srl-pe1-prp.coreb.srl.core.eircom.net [86.43.9.170]
5 46 ms 44 ms 43 ms lag-50.br1.thw.border.eircom.net [86.47.61.254]
6 57 ms 52 ms 44 ms ldn-b1-link.telia.net [62.115.50.13]
7 197 ms 90 ms 80 ms ldn-bb2-link.telia.net [62.115.143.26]
8 47 ms 43 ms 46 ms ldn-b4-link.telia.net [62.115.124.201]
9 50 ms 49 ms 46 ms dtag-ic-323085-ldn-b4.c.telia.net [62.115.50.19]
10 69 ms 90 ms 64 ms 217.239.45.254
11 62 ms 92 ms 64 ms 87.190.232.74
12 60 ms 60 ms 62 ms 195.122.154.3
13 68 ms 71 ms 63 ms 159.100.232.100

fred4

Well, that was working, after logging back in, but now in-game lag and ping dropouts have returned. This is so tiresome. ZOS, I really hope you can correlate these times with your network logs. Do you have an issue with your infrastructure? Does Deutsche Telekom? Are you being DDOSed? This has been going on for weeks, now, on and off.

Mystrius_Archaion

I've been in your shoes with a game before and figured out one main problem with a tracert....

Your lag at the closer steps affects the later steps response time as well because of missing packets that had to be resent.
The "*" on step 6 of your tracert is a time-out of the response on that steps which then affects the following steps as a lost packet, making each response longer farther away in later steps.

Your problem is apparently with eircom.net or telia.net which are steps 5 and 6. Your internet provider appears to be eircom.net. You should contact them with that data for them to investigate.
Hope that helps.


Edit:
Also, those links tend to be hubs in major cities. Load in those cities and through those cities of other traffic will slow down all users and cause some packet loss.

WreckfulAbandon

So... is there hope of fixing latency issues for non IT majors?

Mystrius_Archaion

WreckfulAbandon wrote: »

So... is there hope of fixing latency issues for non IT majors?

Let your internet provider IT majors know about the issue and they will investigate. Use your windows cmdprompt program, which you can search for, to run a tracert to give them data. You can google how to run a "tracert" for easy explanation.

Otherwise, change internet provider or move to a new place. These are the only choices.

fred4

Mystrius_Archaion wrote: »

The "*" on step 6 of your tracert is a time-out of the response on that steps which then affects the following steps as a lost packet, making each response longer farther away in later steps.

I will freely admit that I don't know that much about tracert. I am a programmer who gets dragged into network troubleshooting, at work, when our software gets blamed for a problem.

Can you please consider the following:

(A) In all cases I can continuously ping the 217.239.45.x (hop 10) addresses with no lag and no ping timeouts.

(B) When the in-game problems occur, continuous ping times are fine to 195.122.154.3 and 159.100.232.x, but every 10 seconds, or so, there are packet losses to those addresses, e.g. ping times out.

Are you saying that a problem in Eircom or Telia can explain how I can continuously ping Deutsche Telekom (hop 10), in 60ms to 100ms, but have sporadic failures of continuous pings to Zenimax Germany, performed at the exact same time?

Just to be 100% clear, I am basing my assertion on continuous pings, performed to 3 IP-addresses (hop 10, hop 12, hop 13), over a period of 20 minutes, that are always completely fine and on time for hop 10, but that have outright failures, every 10 seconds, or so, to hop 12 and hop 13. "Every 10 seconds, or so" is an average. The dropouts are erratic and bursty. Could be fine for 20 seconds. Then there's a burst of 1 or more dropouts, interspersed with successful pings.

Mureel

Hi, I CBA to do all that work, but I CAN confirm sudden ping spazziness, lag, suddenly standing still mid raid (awesome as a healer!!), skills and weapon swaps not happening as they should, RIP dps, and even lagging right out of game.

This happened also last night to several members of my guild, all located different places.

Rohamad_Ali

I think the person they hired to set up IP routes is the same person that draws cartoon treasure maps .

Mystrius_Archaion

fred4 wrote: »

Mystrius_Archaion wrote: »

The "*" on step 6 of your tracert is a time-out of the response on that steps which then affects the following steps as a lost packet, making each response longer farther away in later steps.

I will freely admit that I don't know that much about tracert. I am a programmer who gets dragged into network troubleshooting, at work, when our software gets blamed for a problem.

Can you please consider the following:

(A) In all cases I can continuously ping the 217.239.45.x (hop 10) addresses with no lag and no ping timeouts.

(B) When the in-game problems occur, continuous ping times are fine to 195.122.154.3 and 159.100.232.x, but every 10 seconds, or so, there are packet losses to those addresses, e.g. ping times out.

Are you saying that a problem in Eircom or Telia can explain how I can continuously ping Deutsche Telekom (hop 10), in 60ms to 100ms, but have sporadic failures of continuous pings to Zenimax Germany, performed at the exact same time?

Each step is a potential location for packet loss. They are essentially stopping the data and re-transmitting it.
There is a reason a Tracert takes longer than a few seconds to run also. It is actually pinging the time to each hop 3 times, for a result that can be averaged, and timing each hop directly which is why it obviously doesn't just add them all up.

As for the reason I pointed out the time-outs, lower numbers are always better and a time-out means it was either too slow getting a return from that hop or it just stopped somewhere along the way. You have relatively good ping to most of the end hops, but the time-outs you do see are also showing other larger numbers in the same area. The actual data to the final hop, the game servers, is actually rather consistent at a rather low number in the 60-70ms range other than the one time-out which could be explained by the other links in the chain causing the issue.

By the way, many players still connect with no issue to their servers which are using a business grade high bandwidth and much more reliable connection than you own. Potentially, all of these time-outs could be explained by the unreliability of your own connection also.
You have to get everything to each hop and back to record time data for any tracert. That means the connection has to work on your end just to get a value other than a time-out and it has to go through your end twice, and each hop as it goes along twice. It only ever has to reach the server for the game once per test.


Given all that and my own history with tracerts and bad internet routers and providers.
I can say with confidence that the trouble is most likely either on your end with your equipment or internet provider or along the multiple hops along the way, especially those showing slower response/time-outs.

You may want someone to run a tracert from their computer to your modem to see if they see any issues, especially someone who has very few issues connecting to the game. Your internet service provider would actually likely run a tracert to your end in their troubleshooting also.

Mystrius_Archaion

And yes, you will need many tracerts to get a clear picture of what is going on, both when everything is good and when things are bad. You need average pings for each hop.


Edit:
If they're anything like the company I work for, they are constantly running network diagnostics and ping to and from the servers to other locations, other parts of the company, to see if there are any problems.

fred4

@Mystrius_Archaion:

Thank you for the clarifications on how tracert works. Please bear in mind, however, that I only used tracert, so I would know what IP-addresses to continuously ping, in the background, while playing the game. I added the following to my last post, after you quoted it:

"Just to be 100% clear, I am basing my assertion on continuous pings, performed to 3 IP-addresses (hop 10, hop 12, hop 13), over a period of 20 minutes, that are always completely fine and on time for hop 10, but that have outright failures, every 10 seconds, or so, to hop 12 and hop 13. "Every 10 seconds, or so" is an average. The dropouts are erratic and bursty. Could be fine for 20 seconds. Then there's a burst of 1 or more dropouts, interspersed with successful pings."

I have not shown the continuous ping traces in this post. They are basically 3 command-prompts with:

ping -t 217.239.45.238
ping -t 195.122.154.3
ping -t 159.100.232.134

Please explain how there are never any errors continuously pinging 217.239.45.238 (Deutsche Telekom), while, simultaneously, there are errors pinging 195.122.154.3 (Zenimax) and 159.100.232.134 (Zenimax). How can this be a problem with Eircom or Telia? In my view it must be a problem with Zenimax or Deutsche Telekom.

While I value your explanations on tracert, and I agree ZOS infrastructure is far more advanced than my own, I do not see how that has any bearing here.

Mystrius_Archaion

Does Deutsche Telekom host the game or any other real time client you can see real time performance of? Likely, no.
That's the problem with your test.
You may get better results trying to play another game or ping another game's servers for comparison.

There are some apps that allow you to see a real time network graph that may help you. Windows used to have one in system tools, but the newer versions have either taken it away or made it harder to get the same output.

This is a situation where the lowest common denominator rule tends to apply. They run a game for millions of customers, of which likely 100 thousand are easily on at once connecting to the same servers. If the problem is on their end then a majority of those customers will see the same issues at the same time.
More likely, the problem is on your end, or one of the several links between your end and the servers, which wouldn't affect other players, or definitely not as many.

By the way, how old is your computer? your modem/router?
Every electronic device that sees heavy use or high up-time will suffer damage due to heat stress no matter how well ventilated due to the microscopic nature of current electronics.
My current computer is 4 years old. It's suffering some troubles due to overheating over time. I even have the wireless internet controller built into the motherboard crap out every so often and require a "disable then enable" in the system to reinitialize it and get it working from time to time. It has become predictable.

There are several ways this can be on your end or other hops more than just their end. They are a 24hr service provider that manages to keep things up rather well over that time and have dedicated IT staff monitoring things. Your internet provider is as well. Your end likely hasn't had a technician look at it since it was installed.
I would sadly have to say the problem is likely on your end, which could cause every last result you have seen.

And yes, even ZOS and your ISP will have issues from time to time due to load exceeding expectations, or just out of the norm, and potential attacks and hardware failure. If they do have one of those issues though it isn't likely to be common or consistent or they would take the problem equipment down for repairs.

Mystrius_Archaion

And yes, years ago I once blamed the company that ran the game I played for connection issues, and based it on tracert files.
After much work learning more about tracert and contacting my internet provider and testing on other games, I realized it was my end and the distance I was from the servers, causing too many hope, to be the main issue.

I just wanted to share my knowledge so you can skip to the resolution I found, which is changing providers and/or upgrading equipment if possible.

By the way, modems these days should be docsis 3.0 to get the best out of your internet. If yours is labeled docsis 2.0, or worse 1.0, then you would likely beenfit from an upgrade. Wireless routers that are older than 802.11n are likely not ideal also. You will also get more reliability out of a 5ghz wireless band for a router because of less interference from other wireless devices.
I've tried it all and still experience issues on my end from time to time due to the nature of my ISP and my own hardware.

fred4

@Mystrius_Archaion: I am absolutely flabbergasted. I feared such intransigence from ZOS, but you do a better job of fobbing off another player than they could ever do.

There are several ways this can be on your end or other hops more than just their end.

Great, please explain.

They are a 24hr service provider that manages to keep things up rather well over that time and have dedicated IT staff monitoring things. Your internet provider is as well. Your end likely hasn't had a technician look at it since it was installed.

Those are not explanations for this specific scenario. Those are general assertions, based on your experience. It is clear that you work in big business and feel a kinship with ZOS, but what's with the disdain for my home network? This is bordering on a personal insult, since you must have gathered, by now, that I am a programmer of networking software, as I get drafted in to troubleshoot network problems at work. I usually work at the TCP layer, and higher, so would defer to you on matters of network infrastructure, but good God, give me at least a little credit.

I would sadly have to say the problem is likely on your end, which could cause every last result you have seen.

How is it likely? Saying that it could explain every last result doesn't mean it's the first place to look. Not when you have the specific test results I have listed. You continuously fail to provide even a single specific hypothetical scenario that would explain these. Instead you surmise that anything and the kitchen sink could be the problem - and for some reason it must be on my end.

Yes, I have phrased the title of this post as if it's definitely ZOS problem. I still think it's ZOS or Deutsche Telekom, as I have explained in the original post. In case that caused offense, please let me explain that my attitude is "everything is a working theory" until you find the root cause. At the same time, here is my personal pet peeve: Rebooting the system, re-installing Windows, upgrading the network card / PC / router randomly in the hope that any one of these will solve the problem. You learn nothing from that. It's what support people tell you, when they don't have a clue, and they just need to string you along. In effect, that's what you're doing. How on earth can you suggest that heat damage in my PC or router could result in a problem as specific as this? I am not saying it's impossible, but how is that more likely than a problem in Germany?

There's an old joke, which goes like this: It is night. You come upon a drunk who is crawling on the ground underneath a street light. You ask what he is doing. He replies "I am searching for my keys. I lost them somewhere over there.", as he points into the darkness. You ask: "Why are you searching here, then?". He replies: "Because the light is so much better here."

You are asking me to look at my own systems for a variety of reasons, not least that I can do something about them myself. At the same time, the evidence, notwithstanding your assertions to the contrary, points to a problem in Germany. In my view there is nothing wrong with taking that evidence at face value, until you come up with a specific scenario to refute that, but not merely because the light is so much better in your own backyard.

fred4

For further illustration, the following pings coincided with high in-game pings, and lag, today:

Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Request timed out.
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=59ms TTL=53
Reply from 159.100.232.100: bytes=32 time=59ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=60ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=60ms TTL=53
Reply from 159.100.232.100: bytes=32 time=59ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Request timed out.
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=60ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Request timed out.
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=60ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Request timed out.
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=59ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=58ms TTL=53
Reply from 159.100.232.100: bytes=32 time=60ms TTL=53

By comparison, a trace to Deutsche Telekom (hop 10), taken simultaneously, was fine:

Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=64ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=65ms TTL=53
Reply from 217.239.45.254: bytes=32 time=64ms TTL=53
Reply from 217.239.45.254: bytes=32 time=64ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=62ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=65ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=64ms TTL=53
Reply from 217.239.45.254: bytes=32 time=62ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=65ms TTL=53
Reply from 217.239.45.254: bytes=32 time=62ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=64ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=63ms TTL=53
Reply from 217.239.45.254: bytes=32 time=65ms TTL=53
Reply from 217.239.45.254: bytes=32 time=62ms TTL=53
Reply from 217.239.45.254: bytes=32 time=64ms TTL=53

Mystrius_Archaion

1) Your "simultaneous tracert reports" are actually not simultaneous. When the response times are measured in milliseconds the difference is huge between 2 different tests.

2) A better simultaneous test would be 2 tests to the game servers at the same time. That way all hops are considered.

3) I am sorry you interpreted my posts as an attack. I just was pointing out the same thing I also refused to listen to years ago in your same position. I had to learn fro myself that the experts I contacted, both game support and my ISP, were right that it was mostly a problem on my end and my ISP's connection.


Really though, your only option is contacting your ISP. They can check on your end and across the entire network and correct more problems than ZOS can from their specific end. They're essentially just like you with just one end of the chain being their responsibility. As soon as the signal leaves their building it is the responsibility of their ISP, which may even be Deutsche Telekom, to keep it reliable and fix any problems.

If you're unwilling to try a different modem/router then at least contact your ISP citing issues with streaming the game and give them your tracert data to investigate. They can work it across the entire chain, as much as they are responsible for and cooperate with other involved companies.
It's like when a tree takes down a power line and the power company gets out there first and hooks up power they tend to contact the cable and phone companies to have them restring their lines. Power companies are also responsible for replacing power poles so that other companies can attach to them because they don't own the poles.

fred4

Mystrius_Archaion wrote: »

I had to learn fro myself that the experts I contacted, ... game support ..., were right that it was mostly a problem on my end and my ISP's connection.

I have not heard from the experts, at game support, regarding this problem. @ZOS, can you respond, please? I raised this using the in-game help system as well. The ticket number is 170930-002723.

Mystrius_Archaion

fred4 wrote: »

Mystrius_Archaion wrote: »

I had to learn fro myself that the experts I contacted, ... game support ..., were right that it was mostly a problem on my end and my ISP's connection.

I have not heard from the experts, at game support, regarding this problem. @ZOS, can you respond, please? I raised this using the in-game help system as well. The ticket number is 170930-002723.

Good. Check your email for responses. Respond to the copy-pasted automated response for an actual human response.

Mystrius_Archaion

Rohamad_Ali wrote: »

I think the person they hired to set up IP routes is the same person that draws cartoon treasure maps .

It's worse. It's all organic haphazard building of the network based on where people live in greater numbers first and expanding out to lower population areas as demand grew.
Nobody actually planned it out before hand.

fred4

Mystrius_Archaion wrote: »

1) Your "simultaneous tracert reports" are actually not simultaneous. When the response times are measured in milliseconds the difference is huge between 2 different tests.

Please don't tell me that isn't purposely argumentative.

For one thing, not simultaneous tracert. How can I take you seriously when you either don't know the difference between tracert and ping, or your responses are so casual that you can't be bothered with precision?

The pings were simultaneous in the sense that I kept them running for 20 minutes. During that time, there were zero losses to the Deutsche Telekom address, whereas there were losses to the two Zenimax addresses, which coincided with high in-game displayed ping times and laggy gameplay for 10 minutes at a stretch. No, the pings are not technically simultaneous, and the game is using two TCP connections, which firewalls could handle entirely differently. What I am talking about is statistics. During the time I see random in-game lag, and high in-game pings, I also see random dropouts when pinging the Zenimax IP-addresses from a Windows command-prompt. At times when the game works fine, I don't see those dropouts.

These dropouts are to the two last hops, which are Zenimax IP-addresses. While I can't ping the Deutsche Telekom address from the preceding hop, it is nonetheless entirely possible (I would even consider it likely) that the problem occurs at the boundary between Zenimax and Deutsche Telekom, and that Zenimax can, therefore, investigate it.

Mystrius_Archaion

fred4 wrote: »

Mystrius_Archaion wrote: »

1) Your "simultaneous tracert reports" are actually not simultaneous. When the response times are measured in milliseconds the difference is huge between 2 different tests.

Please don't tell me that isn't purposely argumentative.

For one thing, not simultaneous tracert. How can I take you seriously when you either don't know the difference between tracert and ping, or your responses are so casual that you can't be bothered with precision?

Ping and tracert do the exact same thing because of the exact same number results in milliseconds. It wasn't meant to be an argument but a simple fact of all electronics. Electrons move at nearly light speed. You're never going to get a simultaneous result or it affected by the exact same conditions.
This is the very effect of the "uncertainty princple". You can never know the exact position and speed and direction of a subatomic particle. The very act of measuring such also changes those values because of how we measure.


I'm done with this. You're either going to do what works and contact your ISP and deal with ZOS support through email, off the forums, or you're going to aggravate yourself to the point of hating the game completely and maybe self-destructing your account.
I've left a game over performance before, but the difference is I knew it was the game's fault even if my own end triggered the problem. It was Champions Online which had "infinite rubber-banding" due to stupid position coding.

theivorykitty

fred4 wrote: »

I am a programmer who gets dragged into network troubleshooting, at work, when our software gets blamed for a problem.

Real programmer confirmed.

But, seriously, my father, husband and I are all IT professionals and are all having major connection issues from different houses and offices. Logging in fails 3/5 times, and zoning often kicks us out. We know our equipment is up to par, and we know the connection issues are only related to ESO, specifically. This has been happening since the last patch. It sucks.

Glad to see another post about it, because no one answered mine except to suggest that we all need to restart our routers (which, of course, we've already tried anyway as #1 possible fix).

fred4

theivorykitty wrote: »

This has been happening since the last patch. It sucks.

Yes, it's only happening over the last weeks, for me. I can't remember excactly which patch, though.

xRIVALENx

If you happen to be routed from Dublin Level3 to Frankfurt Level3 this may be a contributing factor to your latency, seeing upwards of 2400ms spikes.

fred4

That's an intriguing piece of info. Care to elaborate on how you know or why this would be?

xRIVALENx

Not sure why it is happening but I did run a ping plot all morning viewing the hops between Dublin and Frankfurt simply because I wanted to see the latency from 159.100.232.134 for a connection out of Ireland.

xRIVALENx

Rohamad_Ali wrote: »

I think the person they hired to set up IP routes is the same person that draws cartoon treasure maps .

Surprisingly this isn't far off when viewing the level 3 network map around the UK.