What's new

Looking for help buying an sqm Router with performance to almost 500mbit connection

  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

The 30msec is several hops out from them, which implies their ISP is fine, at least to a certain point. @Furious - ping test/traceroute a few other sites, specifically ones within your country, that will give a better view, if they all hit that 30msec hop, then your ISP may just have a lousy route to the internet backbone, or some other routing problem on their network.

Those 4 first hops seem to be from the same company - OTE in Greece. How a home connection reaches the backbone and cloud providers has everything to do with the ISP and not a single home router can help here. Maybe other ISPs have the same issue in that location, but it's worth a look...

My neighbour pings Google under 10 ms, while I get 15-30 ms. The more my packets travel, the bigger the latency gap between us....and obviously, we do not share the same ISP.

 
Last edited:
Those 4 first hops seem to be from the same company - OTE in Greece. How a home connection reaches the backbone and cloud providers has everything to do with the ISP and not a single home router can help here. Maybe other ISPs have the same issue in that location, but it's worth a look...

My neighbour pings Google under 10 ms, while I get 15-30 ms. The more my packets travel, the bigger the latency gap between us....and obviously, we do not share the same ISP.


Yes, like I said their connection to the ISP is fine, quite low latency in fact. Your path to google may follow your ISP for several hops, it may jump onto a backbone. For another site it may be totally different. All depends on where you are located, how many connections to the backbone your ISP has, how their routing is designed, etc. As you say no router or QOS is going to do anything about the latency that is more than a hop or two away from you.
 
Oldie but goldie:

dlakelan's advice is very good, if you need a 500mbps or above link.

Another post in that topic on the OpenWrt has contrarian advice: instead of spending more on your router, drop your ISP speed - a modest router at 100mbps with good SQM (10-15 msec latency) will outperform a 500mbps link that has 100 msec of bufferbloat. https://forum.openwrt.org/t/so-you-...er-and-need-a-router-read-this-first/90305/40
 
Over the last 30 days I have uprated my FTTP (via BT) connection from 100/20 to 500/70. Changed my router from RT-AX88u to Synology RT6600ax. I monitor my connection to the Internet via a Sam Knows white box. I have run the graphs on UDP latency for the last 30 days. My Latency has stayed the same at 80 ms throughout. Latency is measured frequently throughout the day and not just "spot checked" Have you heard about PingPlotter ? It show the "delay" when pinging a destination and the hops inbetween. It can show you if the route to the gaming server is changing which can dramitiacally affect your experience, From my experience there is virtually nothing you can do to controll Latency/Delay over your complete path to the gaming server.You can prioritise traffic locally but once it arrives at your ISP all the Priotiy flags are reset. Perhaps move closer is the only viable option, although not very practical.
 
Here is a simplified way of looking at it:
-No QOS and not hitting any network limitation - Lowest latency
-QOS while hitting a network limitation (like saturating your speed) - Medium latency and moderate packet loss (higher for some devices than others, that's basically how QOS works, deciding which devices to drop packets from).
-No QOS while hitting a network limitation - highest latency and highest packet loss
How about "Yes, QoS and not hitting any network limitation*" though?

* which is where the devil already lies in the details as the usually stated utilization of a network link is an oversimplification, i.e. average over time only. If the time span is short enough, the momentary usage will always be 100%. Actually it's a binary duty cycle of bursts of 0% and 100% usage, taking changes in the line coding or modulation aside.

With that premise in mind, I wonder whether QoS isn't also to one's advantage, at least in theory, even when the total utilization isn't at the maximum (measured over a long enough time span). Shouldn't it be preferable to send more urgent packets first as they would still take a slightly longer time otherwise due to the always finite achievable transfer rate?
 

Similar threads

Support SNBForums w/ Amazon

If you'd like to support SNBForums, just use this link and buy anything on Amazon. Thanks!

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!

Staff online

Top