What is a modem+router good enough for online gaming?
I recently got an Ethernet cable in the hopes of making my online gaming more responsive, but to my dismay it made little difference in latency measure on the Xbox Series S. It merely dropped from 146ms to 143ms.
I use the modem+router provided by the ISP, a Sagemcom Fast 5655v2. According to preliminary research, the ISP blocks any alterations so I would have to jailbreak the device to explore other solutions. I’m open for suggestions in that regard too! I’d like to know if I can determine if the problem is on the router or the ISP.
On your suggestions please consider that my country’s currency is worth less than one fifth of the US dollar, so I’m not looking for anything even remotely close to the best setup possible, but merely a significant improvement. Anything above 50 US dollars is already too much for me.
So, with that in mind, what do you recommend?
At first glance, I'm not sure there's much to be gained by swapping out the modem/router. Unless they're very very old or defective in some way, I'm not sure that newer tech would improve your ping time at all. That usually has more to do with the infrastructure outside your control, unfortunately.
Sorry I don't have any helpful suggestions to add.
Having your own hardware will always be better than whatever junk the ISP is providing. You definitely do see an increase in ping and speed. Most ISP hardware is extremely out of date and cheaply made, never mind the fact you waste a ton of money renting the hardware. A single year of Comcast rental fees is enough to buy your own hardware that will be better.
I would never advocate users rent the hardware from their ISP purely because it is a poor value proposition.
However most major ISPs provision reasonable modem/router combination units that are perfectly fine for the majority of people. The problems users might run into with them are usually lacking feature sets or clunky software, but rarely poor performance.
OP should buy their own modem and router, but they should not expect this to make a noticeable improvement in online game performance.
Can't say I agree. ISPs are starting to put out some better stuff. And since they configure it it's often better for customers
Usually there's an option to buy the ISP's internet hardware directly-- I know Comcast offers/offered the XB3/4/5 for actual purchase (I don't know if the XB6/XB7 continues this trend, since I both stopped working and being a customer of theirs before they came out), and Centurylink gives an option to buy one's modem at install time from them (plus there's a decent secondary market for their equipment on eBay/Amazon and the like). If that's a concern, that's the route to go down.
Plus at least on the Comcast side, I could troubleshoot purchased modems that were Comcast branded the exact same as I could leased ones (do wireless network changes remotely, run our diagnostics, and so on), unlike the third party purchased where I just had the standard DOCSIS troubleshooting tools. If someone has less tech-savvy family or friends they've talked into purchasing their own modem, this is another thing that can help you from becoming free tech support for them. ;)
That is most certainly useful, since you may have prevented me from making an unnecessary purchase. Thanks!
If it's a cable internet connection, odds are around 100% that it's slow because they've spliced too many people into the last mile. Somewhere not far from you is a large metal box that has far too many Y-adapters and probably had to be shouldered closed. Sometimes you can call the ISP and they'll send someone out to check it. Just tell the guy to move your wire up the stack, that's about all they can do.
However, you can also trace your cable inside the house. Make sure the router is in front of the cable splices that go to any televisions rather than behind any Y-adapters. Also it's possible the adapters you have inside are just old and messing up the signal. I've seen this happen many times. Easiest test is to just run the cable itself straight into the router with no other tv/devices attached at all and see if it helps.
Keep in mind if you trace the cable you are liable to find a rat's nest of adapters in there somewhere, likely in some closet behind a panel. Just because you can't see the adapters doesn't mean they aren't there. If you are in a rental unit it's worth a look - all those clean wall plugs meet up somewhere. The cable techs have plenty of replacement cables and adapters in their vehicles, they'll be happy to do a trace/debug/swap if you place a call. They also have signal testing devices to make short work of this stuff.
For comparison, I have a 30ms ping on a cable connection. I'm also out in the boonies with all of maybe 30 neighbors sharing this segment and my router is the first device here. Rental units and/or cul-de-sacs and other... challenging network configurations can stack the number into the hundreds, and DOCSIS does not scale well. It still works (and perfectly for video) but it becomes sluggish for TCP/IP. If you talk to the cable guy, ask him what version your segment is using and if they plan to bump it to 4.0 anytime soon.
Can you use a tool like WinMTR (or just normal old MTR) to see where the latency is coming from? You'll want to pick a destination that matches the servers you'll be playing games on. My guess is that you'll see the latency picks up far outside of the hops you control. Even so, there's a chance you can force an alternate route to get better latency.
From your computer, you could do a traceroute to find out where the bottleneck is. Like others have mentioned, if you are hard wired, it's most likely not between you and the wall. You'll have to find the IP or hostname of the server your XBox uses for online activities. I found this and it might help https://pastebin.com/UjCKCejT
First things first, be more exact about what you're measuring. That test on the Xbox might be measuring ping time to Microsoft HQ in Redmond, or whatever.
This speed test from Cloudflare may be helpful, it measures ping time as well as showing the location of where your latency is being measured from.
If you can, run that test on both a wired and wireless connection, and it'll give you an idea of how much your wireless network contributes to your latency. For example, on my Ethernet-connected desktop, I have 2.09ms latency and 2.63ms jitter (I have very fast gigabit internet, and Cloudflare's server here in Seattle is ~10 km away from me).
But, on my wifi connection, it shows 3ms latency and 15ms jitter - showing that even though the average is almost the same, the wifi connection's latency is much more unpredictable, probably because I'm in an apartment building with a very congested wifi radio spectrum.
Another useful thing you can look at, the VM provider I use, Vultr, provides public ping-able endpoints in each of their datacenters.
If I ping their local Seattle datacenter I see a glorious 1ms ping.
If I ping their Sydney datacenter, it's around 170ms. That's even with my gigabit internet - as fast as it is locally, it can't escape the speed of light. Great circle distance from Seattle to Sydney is ~12,500km, which is 41 milli-lightseconds. That makes 170ms latency really impressive - all the network gear, and the indirect cable routing, adds ~130 ms of latency, and meanwhile the fucking speed of light is responsible for the other ~40 ms.
Vultr doesn't have a South America region, but if you compare your ping times to their datacenters in Australia / Asia / Europe / US, you should be able to figure out how much latency is local to your network and how much is inherent in your traffic trying to escape Brazil.
Realisticly, that kind of latency is more likely due to your ISP or the server you are connecting to.
If you're in USA East connecting to USA West, that would easily be 70 to 80 ms just due to distance.
Infrastructure plays a large role. I'm in the USA, and I have a choice of cable or fibre. Cable's older infrastructure, between more conversions and breaking cables gave me 40 ms lag to same server I get 12ms on now.
You should try gathering some data. Fast.com and other speed tests, try to do them different times of day and see what changes are like. If you're consistently getting high pings and/or very low throughput, you should bug ISP support until you get someone to fix it.
It might be interesting to see what the latency is to an Internet service that’s closer to you, for comparison. (If it’s closer then the traceroute output should have fewer hops.)
Do you have another ISP option available (preferably one using different physical lines from your existing one)? This almost sounds like they changed routing or peering and your connection is just taking a long hop route. Unfortunately, that 1) changes all the time and 2) isn't something that gets changed based on individual customer complaints.
One thing you could do, if you have a VPN accessible to use, is try connecting to it and check your latency to it post-connection-- if the VPN provider has different peering arrangements, you could see lower latency if it's a routing issue instead of any physical plant problems. If you get noticeably lower latency on the VPN, that points to ISP peering or potentially issues upstream of them, and swapping ISPs (or just using the VPN full time) might be the best bet.
Do Brazilian ones have the constant new customer promo offers like they do here in the US? Could end up saving you some cash in the long run, depending.