High Ping to Portland from Medford, OR

Options
gardenfrog99
gardenfrog99 Posts: 7 Spectator
edited May 2023 in Gaming & Gaming Related

I should get 20ms ping to Portland, but it is 52ms. I ran a trace route and everything is fine until the last server node. My ping to the Medford server is normal at 8ms. But at some point in the line my ping jumps up above normal on its way to Portland,

lag-802.prr01sttlwa.netops.charter.com [96.34.3.39]

At this node my ping jumps up to 50ms. Just last night my ping was fine and this node was at 24ms. This has been happening for a few days. This happens when I am just connected only to the modem so it is not due to router.

I was having this issue of high ping to Portland and so I went and exchanged my modem. Which for a few days fixed my problem. I was getting 20 ms pings to portland but then just today that stopped. now it is back to 52ms.

Another interesting fact is even though when I had 20ms pings to portland via speedtest.net, my game fortnite, their servers are located close to portland, still had high pings...but when i would run a trace route for their servers the trace route would come back with normal 25ms pings but in game I was experiencing 50-60ms. But this is just less important. As doing a simple speedtest.net to portland with modem only results in 52ms which is about 30ms high compared to what I have always got before which is 21ms.

Best Answer

  • RAIST5150
    RAIST5150 Posts: 918 Contributor
    edited April 2023 Answer ✓
    Options

    Well.. it is a popular online game. By nature it is going to be prone to do weird things like this.

    You are likely being impacted by common inconsistencies that you have managed to avoid for quite an unusually long time.

    A portion of that can actually come down to issues with matchmaking in many games though, not necessarily the fault of your ISP. Ties into something you alluded to (about players having half the ping of others in the instance).

    Unfortunately, the numbers you are demonstrating are basically within what has become an accepted range on a larger scale. To be frank, cable internet is not really a good choice if one aspires to play at a level even remotely equivalent to an "e-sport" tier of competition. As mentioned previously... median latency of 26ms.

    With decent lag compensation in play at up to a 30hz refresh (roughly 33ms time slices), people should (in theory) be fine up to about 60-ish ms pings... able to compensate to some degree up to around the 80-90 mark...

    ...provided things are consistent.

    Throw people into such a lobby that are able to stay highly in synch with the hosts updates via 32ms or better RTT, and it will cause problems.

    It is something MANY franchises have wrangled with for years... how to best weigh each factor when matching up players.

    Fortnight is one of those games that relies heavily on cloud servicing to run the instances. The issue with such design is the assigned hosts can move around. In some games it can even migrate mid game. Sometimes such cloud servicing simply can't balance everyone's latency profile well.

    The fact that the profile of the potential player pool can shift in weird ways at any time for a host of reasons can compound things further. But it is pretty common to see things get problematic when players move to new ranks that throw them into smaller pools.

    If this indeed is part of the problem, you may see a pattern of sorts if you actively track and locate your connections for a while. It is the kind of thing that drove the likes of NetDuma and WTFAST into popularity. Even when on fiber, some feel it is worth the investment to use such things. Then there are those that have just had to learn to live with unruly pings in their games until things eventually smoothen out... adjusting to the frame hitching, leading their targets, etc.

    ¯\_(ツ)_/¯

Answers

  • Lyn_T
    Lyn_T Posts: 423 ✅ Verified Employee Moderator
    Options

    Good evening @gardenfrog99 and Welcome to our Community Forums.

    Thank you for reaching out to us in regards to your concerns with your ping rate. I was able to locate your account using your registration information. Generally the ping rates you are reporting are within the normal range. Is this happening on all sites you visit or is this happening on just with Fortnite? It looks like you have a home visit scheduled to have this investigated further. Is that correct? -Lyn

  • gardenfrog99
    gardenfrog99 Posts: 7 Spectator
    Options

    First, these pings are not normal for me. I have had service for 3 years and have always had 20ms ping to Portland. I just had 20ms ping yesterday. It is not just fortnite. When I run a speed test to portland as I stated above it is 52ms. So there is something going on. For some reason the ping is jumping up to 52ms and I have no clue why. I went and replaced the modem and for a few days I had 20ms ping to portland. Then I wake up today and it is back up to 52ms which effects all programs including fortnite.

    Here is the trace route... you can see at node 6 is when the high ping hits in. I had just done a trace route last night and there was no increase in ping. So to say these are normal ping times is incorrect as for the distance to portland i get 20ms give or take for years and suddenly on april 7th it shot up and i have tried multiple things to fix this. i am technically proficient and have done almost everything. and yes i have a tech coming out tomorrow to maybe run some diagnostic or at least get this info back to higher level engineers that focus on trace routes. it makes no sense that my ping would fluctuate 30ms like this...but once it gets stuck at 52 nothing brings it down really except replacing the modem.

    1   *    *    *   Request timed out.

     2  16 ms   7 ms   7 ms lag-51.dtr03mdfdor.netops.charter.com [96.34.105.232]

     3   9 ms   8 ms   7 ms lag-201.crr02mdfdor.netops.charter.com [96.34.110.28]

     4  10 ms   7 ms   7 ms lag-808-10.bbr01mdfdor.netops.charter.com [96.34.108.168]

     5  16 ms  23 ms  23 ms lag-4.bbr02sttlwa.netops.charter.com [96.34.0.56]

     6  48 ms  48 ms  49 ms lag-802.prr01sttlwa.netops.charter.com [96.34.3.39]

     7   *    *    *   Request timed out.

     8  49 ms  51 ms  91 ms 52.95.52.195

  • gardenfrog99
    gardenfrog99 Posts: 7 Spectator
    edited April 2023
    Options

    so for some reason i am now getting 21ms times to portland. but one interesting thing is that in fortnite i have always got 20-30 ping(for 3 years straight). since april 7th it has went up to 50-60 even when a speedtest to portland shows 21ms where fortnite servers are close to. but the weird thing is when i connect to my vpn in seattle my ping is lowered to 35-40. thus for some reason spectrum changed something that made my ping go up and is routing my interenet traffic somewhere out of the way or has a faulty server...i am not sure as this is beyond my scope and given data. so if you are somehow able to contact high level engineers about this issue i would be very thankful.

  • RAIST5150
    RAIST5150 Posts: 918 Contributor
    edited April 2023
    Options

    Routes can change with new announcements. Sort of an autopilot thing that happens within the various routing protocols in play across the interconnected networks that make up the world wide web (BGP, IGP, etc.).

    As such, there is a perceived "normal" range that is expected to be seen. I suspect that is what was meant by that comment about the response times being normal.

    Keep in mind the latency of HFC's has traditionally been a bit higher then pure fiber by comparison, but it has gotten much better in recent years. Haven't looked at the FCC's latest report yet, but the Spectrum policy page notes The median for Charter as 26 ms across all markets:

    (Note this is a very specific test, like MPG ratings on cars... YMMV. The figures are more for comparing against how others faired against the same test parameters).

    It used to be considerably higher when we had less bandwidth available.

  • gardenfrog99
    gardenfrog99 Posts: 7 Spectator
    Options

    well for over 3 years I had 20 ping to portland and now it is 52. essentially my vpn routes the signal all the way to seattle and then back to portland(really eastern oregon) faster in fortnite than simply using the isp route. I think maybe there is local network congestion by a neighbor or something which would explain why sometimes it is 52 and other times 20...but also could be something else. it is not consistent but is consistent in that it is either 21ms or 52ms giver or take 1-2ms. I have never had that high of ping until last friday so would like to get some data from a proper engineer on what may be happening if possible.

  • RAIST5150
    RAIST5150 Posts: 918 Contributor
    edited April 2023
    Options

    Considering you may miss the same number of frames (or more) in the blink of an eye (around 100ms)... not sure if it will get very far. The variance you are reporting is typically handled with basic lag compensation in games.

    If you were tracking elevated packet loss, or more extreme ping times.... different story. But considering 60fps frame times come in at 16.6666...ms (not sure if Fortnight's tick rate can even keep up with that... may only be 30hz), not sure how much traction this will get..

  • gardenfrog99
    gardenfrog99 Posts: 7 Spectator
    edited April 2023
    Options

    oh i severely doubt it will get traction...but you do not play fortnite obviously. there is a huge difference between 20 and 50 ping. players with low ping have a huge advantage at the higher level of the game. there is even a huge difference every 20 ms ... there are no games like fortnite and very few players can play competively over 60 ping and most have 0-20 ping. once you get up to 50-60 ping you are at a major disadvantage. fortnite can be decided in half a blink of the eye.....it is the reason i play it and have never played any other game. it is a great exercise for the brain. i just think spectrum shou8ld be embarrassed that their routing is slower than a vpn when i live in southern oregon but a signal can go all the way to seattle and back quicker than they send it straight to portland. i met another person with spectrum having the same problem.

  • gardenfrog99
    gardenfrog99 Posts: 7 Spectator
    Options

    who knows what is causing this. when i use a vpn to seattle my ping is 20ms faster. so it leads me to believe it is the isp routing differently as before when using a vpn my ping would be slower do to the obvious fact that unless your isp is routing stuff strangely then a vpn will make things slower always. so saying this is normal etc is in fact ignoring the isp routing issue which seems to be the most likely of causes. i have another spectrum player in m y area and their ping jumped too so the chance of it being a spectrum routing issue is by far the highest. i appreciate your time but what i need is data regarding the route my signal takes to its destination and looking into possible making changes to that route for those of us gamers in my area. but i seriously doubt i will get the help i need. luckily i have a vpn which lowers my ping but still causes lag compared to no vpn gaming i have had for over 3 years. your other conclusion that i have been lucky is blatantly very unlikely.

  • RAIST5150
    RAIST5150 Posts: 918 Contributor
    Options

    Already touched on that briefly the other day...

    "Routes can change with new announcements. Sort of an autopilot thing that happens within the various routing protocols in play across the interconnected networks that make up the world wide web (BGP, IGP, etc.)."

    Perhaps more focus is warranted here...

    As long as the next hops returned in a query continue to function within a service's established parameters (not uncommon to see metrics in the ballpark of 100 or so, maybe towards 150--depends on their policies), they usually don't intervene. Something like a shift from around 100 to 250 may warrant it, but they may opt to hold off a bit for the system to autocorrect first--unless of course it is REALLY bad or they are just getting hammered about it.

    When you run through certain VPN's or Proxies, you may directly impact the next hop lookup when the query is run for the BGP/IGP data. If when you exit there is a big enough shift in geolocation or you simply wind up on a new ASN up front, it may dramatically transform things... almost as if you switched ISP's. In a sense, that is what happened--the next query comesfrom a different location, lilkely against a different route server than your ISP's.

    Sometimes you come out ahead, sometimes it is worse.

    Think of it like choices when planning a flight. Which airline you pick and which airport can alter more than just departure and arrival times... in flight movie and refreshment/ meal options (if any), as well as your connecting flights (or even eliminate the need for any). It can dramatically alter the entire experience.


    This route data is routinely updated in batches in the background, but it is not done on any form of a standardized interval. You may see some records change every few weeks, others may age 100 days or more.

    Recently spotted a range of new announcements hit right before I queried a route server for someone hosting on the west coast. Was looking at routes from a specific Spectrum exchange on the east coast to a specific host on a west coast Comcast IP. Saw new announcements from level 3 and Akamai, among others. The one record that stuck out to me though was where I defaulted, a familiar sjc address. That leg from Atlanta has been a metric of 100 for a looong time... talking back to when SE moved XIV hosting from Montreal to California. It was now showing 129. Didn't see another route with a lower metric to supercede it. I was still hitting the last exchange before the target server for around 90ms or less depending on load though, so no change for me--but it may impact others depending on what is going on in their "first" and "last" mile legs.

    The point of mentioning this is that apparently something changed some time ago out there causing it to register a higher latency, so a higher metric was catalogued for that route to aid future route decisions. This was not likely done manually by Spectrum or Charter (was actually querying HE's route server). It appeared to be a routine batch of updating by the "system" that runs in the background to keep the internet moving along.

    This happens all the time.

    Network conditions are in a perpetual state of flux... why so much of the management is automated, as it would require way too much micromanaging to do it by hand. Also why people pay for a service like WTFast that curates specific routes for specific known game hosting servers. They are effectively doing the same thing as any other VPN or Proxy service... they just have a specific focus on how they reach known gaming hosts.


    This may just be a short term situation that eventually reverts back on its own (in the context of the internet, this can be longer than some may think). It could become a more permanent change... who knows.

    Not really much point digging into the "why" such a minor adjustment of 29ms was made there. Could be the result of infrastructure changes, or just some weird congestion issue that reared it's ugly head and nothing has warranted reverting it back to 100... yet.

    The key thing is that what you are experiencing will generally be considered a "normal" thing here... as it is in fact a "normal" and "expected" thing to happen on occasion.

    There is no real guaranteed QoS agreement to enforce here like you may get with a more dedicated service (like a T1). As small of a variance as this is (in the grander scheme of things), it likely will not draw much attention and get changed manually... unless it gets MUCH worse than a shift to 50-60ms RTT.

  • gardenfrog99
    gardenfrog99 Posts: 7 Spectator
    Options

    yeah, i am definitely realizing this issue is not resolvable except possibly with time. i am fine using nordvpn for now since my ping with the vpn is 20-40 which is way more playable than 50-60. 50-60 is simply not fun after being used to 25. fortnite is unique as when building and editing you are making adjustments where if you have 60 ping you just can't do...look up triple edit fortnite for instance. over all a good learning experience for me about how the internet works and i appreciate your time and wish you the best. looking into wtfast and it seems kind of like the many scam like companies out there offering better gaming ping. but it does have the best reviews...i tried a FEW DIFFERENT VPN'S TO SEE IF THEY COULD BEAT NORD BUT NOT REALLY EVEN WHEN THEY HAVE A SERVER IN PORTLAND CLOSER TO FORTNITE'S SERVERS...oops caps...and yes i know fortnite uses amazon servers. which are actually located out close to the dalles oregon and san jose though they are very secretive about the actual locations. i wish i could get fiber in my area but only one company does it and they are building neighborhoods at a time and prob wont get to me for years by which i may be done with gaming. i just find it helps keep my brain active and is a good mental exercise. another thing is that i get 18 ping to seattle and 20 to portland when portland is a few hundred miles closer. very interesting how internet works...

  • RAIST5150
    RAIST5150 Posts: 918 Contributor
    Options

    Out of curiosity, I start poking around some of their EC2 compute stuff... one of the few areas where you can straight up get legit IP's and subnet data.

    Wondering if that 52.95.52.195 address you posted the other day is even very close to you. Wasn't getting any FQDN coming back when I dug on that IP. What little geodata I found was all over the map. Weird one indeed.

    Here are some legit addresses from their EC2 reachability page for us-west-2 Oregon (got that naming scheme from cloudping.info page). Geodata was assigning them to Boardman, and their times came in the 100-120 range for me from South Carolina.

    52.43.255.254

    52.88.0.2

    54.68.63.252

    54.70.0.12

    This is the one Epic says to ping for testing your region (though it appears to be for SanJose):

    ping-naw.ds.on.epicgames.com

  • Jasper1675
    Jasper1675 Posts: 1 Newcomer
    edited April 2023
    Options

    Raist5150,

    I wanted to ask some questions on this exact topic since the original poster and I have the same issue. There is another layer to problem we are having. We both want to play some games on Fortnite's tournament server hosted by AWS in Dallas. The 20-30ms ping to 50ms between the two west servers is annoying but I doubt will get any traction.

    The issue seems to be Spectrum is routing almost all of our data to Seattle WA then back to us in Medford. This adds about 35 -40 ping for this round trip. I'm stuck with Spectrum at my house but in other places I have other internet providers I can test who are located in Medford. The issue with Dallas is this extra hop seems takes the ping from low 60s to 105-115. 60s is playable 105-115 is not really functional in Frotnite. I have run numerous pings and to Dallas from the Medford area and 60ms is the norm if you aren't on Spectrum, Spectrum is 105-115.

    I have run a couple trace routes and almost without fail I go to Seattle First on Spectrum. One time I ran a trace route to AWS in Virginia and I went to San Jose first instead of Seattle.

    If I could reroute my data south out of Medford instead of Seattle my issue would be resolved. Do you know of anyway to do this?

  • RAIST5150
    RAIST5150 Posts: 918 Contributor
    Options

    Short of "homing" your connection elsewhere, not much the end user can do directly... may have to wait it out.

    Perhaps a lease renewal can get you to another network segment and that dramatically changes the next hop election.

    So long as you stay as you are now, it may simply not change until either the the route server updates to a cleaner route, or whatever congestion or other issue that is forcing that election to take a roundabout detour clears up.

    In the short term, a proxy or VPN may do the trick while <hopefully> providing some options to choose from.


    We ran into issues like this back when FFXIV was in betas and through the initial launch of the game. Routing through specific peers were always solid until things got congested and people would get flipped to alternate peers... sometimes it would just fall apart for some markets. They simply were NOT prepared to handle all that traffic from NA and EU to get crammed into Montreal all at once.

    Our market here was still with TWC back then and had a means of engaging with tier3 via the community team (in some cases even directly by callback). It took a lot of effort and persistence on the players gathering data and forwarding it to different entities for analysis (Net Neutrality was still a hot ticket... admins seemed more accessible in general back then). Then one day, after about 6 months, the Eastern seaboard got cleaned up as a whole.

    Down the line SE moved the hosting to California though... and it all went to crap again.


    While we were all wrangling with Montreal, some used VPN's to get away from Level3 Atlanta and certain ALTER.NET segments (Verizon related) in the NY area that had been identifed as problematic. Some would factory reset their modems or just power down overnight in the hopes of getting on a different subnet and such. All the while forwarding traces, pathpings, BGP route data, etc. to their ISP admins. Some even tried forwarding the same data to Level3, TATA, and Cogentco (relevant peers to Montreal along the east coast at the time) as well as SE Support.

    To this day, we still do not know if WE actually affected the changes or not... took so long to happen. But the latency DID eventually get cut roughly in half to the specific ASN we had identified to be in use by SE Montreal.

    There were still issues on SE's end of course, not to mention any potential snags that could crop up on residential lines locally. But nixing 60-80ms off the RTT from the southern states to Montreal was a huge game changer. It just took forever to happen.


    That is the frustrating thing about our infrastructure in general. Things can tank at the drop of a hat... but recovery can take forever and a day sometimes. The gateway protocols will kick in and shunt traffic to another route to keep data flowing and all. But those alternate routes often have very bad metrics... why they were not in mainstream use in the first place. And we get stuck with it until whatever issue that cropped up is resolved or new and better routes get announced/assigned.


    If we had contributed anything useful towards resolving the Montreal problems, it was likely providing verifiable endpoint data for the service in question: IP information that could be confirmed to be used regularly by the game, and could be actively monitored/tested against to collect loss and jitter data consistently. Providing data from within the game client and such does not help much if it cannot be tested externally (have to play the game to test it? Lol).

    If they cannot see the issue occurring live, they cannot tweak/test things properly to confirm if the problem persists or got resolved.

This discussion has been closed.