What is bufferbloat and how do you notice it?
In 2010, Jim Gettys, a veteran computer programmer who currently works at Google, was at home uploading a large file to his work server. His kids came into his study and said, “Daddy, the internet is slow today.” Wondering how his upload activity could be affecting downloads by his kids, he began to investigate.
By experimenting with pings and various levels of load on his internet connection, he discovered that latencies were often four to 10 times larger than what should have been expected. He termed the phenomenon “bufferbloat.” His conclusion was that critical data packets were trapped in buffers that were excessively large.
From the time Gettys made his observation and began to publicise it, researchers from companies such as Cisco and Google, major research universities and standards groups like the IETF have been investigating, testing, and writing about bufferbloat.
While the term ‘bufferbloat’ will be rather meaningless to most people, it’s devastating effects are very widely noticeable. With bufferfloat, other devices or applications running on your network can disproportionately wreak havoc with delay-sensitive applications you may be using, such as online gaming, web browsing and voice telephony (e.g. Skype and Zoom).
Those with high-speed connections are not immune. With an increasing use of video streaming from YouTube and Netflix etc. and an increasing number of Internet-connected devices in a typical home, bufferbloat will continue to get worse.
According to expert Jim Gettys, “Bufferbloat is responsible for much of the poor performance seen on the Internet today”. So, if you’re struggling to understand why your upgrade to superfast or ultrafast broadband connection has not given the performance boost you had hoped for, or expected, bufferbloat may well be the root cause. This is because critical small data packets that have to be transferred in a timely manner for applications to work well (e.g. VoIP packets, DNS look-ups and TCP ACK acknowledgements) can be trapped in the buffers of network devices behind much larger packets associated with streamed video and file transfers. These delays can wreak havoc in a number of highly-noticeable ways:
online games can suffer from unacceptable lag
web browsing can become sluggish, and less responsive
delay-sensitive applications such as video and audio telephony (e.g. Skype) can become severely degraded (e.g. with video freezing, audio drop-outs and poor quality)
streaming video services pause and break up
Many people assume that the solution to their broadband issues is to increase the speed of their connection but high-speed connections can suffer dramatically from bufferbloat too. If you sort out your bufferbloat problem, you may be amazed how good your current broadband service is.
Bufferbloat is getting worse in many households, fueled by the demand of video streaming
The widespread availability of superfast broadband in the UK, with 95% of homes now able to get download speeds of 30Mbps or above, is helping to fuel the adoption of a range of broadband-delivered, bandwidth-intensive services to an ever-increasing number of Internet-enabled devices in the home.
These trends are inevitably making bufferbloat much worse. Many people who have upgraded to superfast (or ultrafast) broadband in the hope that simply increasing broadband speed would solve their performance issue may be bitterly disappointed.
Internet forums are full of online gamers who are struggling to achieve lag-free gaming while others within the household use streamed video-on-demand services such as Netflix and YouTube.
How to measure bufferbloat on your network
Most online speeds tests that measure latency/ping such as Ookla (Speedtest) measure these when nothing is being downloaded or uploaded, when the broadband connection is not fully loaded. The most comprehensive speed test for the measurement of bufferbloat is provided by DSL Reports. You can measure your bufferbloat here:
The DSL Reports speed test continually checks latency from the start to the finish of the test, not just at the beginning (like most other speed tests). If the latency rises substantially during the download or upload section, this indicates a high degree of bufferbloat when the connection is fully loaded.
DSL Reports provides a score for bufferbloat – between A+ and F. Grades are awarded according to the increase in round trip delay (averaged across downlink and uplink) when the connection is fully utilised. Grade A+ is awarded to connections where the average increase in latency is less than only 5ms, which is an outstanding result.
Based on data from DSL Reports, most connections suffer an increased latency of 60ms or above, allowing you to judge your own connection. To get the equivalent of an A+ rating from DSL Reports, you need to achieve an increased latency of less than 5ms. This is achievable on Guardian Security App Routers if you spend a little time and have the patience to tweak the Router settings.
In general, the most effective way to combat bufferbloat is to implement Quality Of Service (QOS) in your router. Guardian Security App routers will be running the most effective SQM algorithms that specifically address bufferbloat, such as FQ-CoDel or the more advanced CAKE (Common Applications Kept Advanced). These control how traffic is managed on your network and ensure that the small packets that are so critical to latency-sensitive applications are not delayed in buffers behind packets for video-on-demand, for example.
Guardian Security App QOS Basic Setup
Always backup your router before making any changes
Log into your Guardian Security App Router
Select the NAT/QoS tab and then the QoS sub-tab.
Click "Enable"
If your Guardian Security App Router is to be setup for multiple services such as Voice, Video and Downloads, choose HTB. If not, a better throughput may be achieved with HFSC.
Set your download and upload speeds. You can use a speed test like the one at dslreports.com/speedtest to check your actual connection speed. Enter no higher than 95% of the values you measured into the proper fields.
There is no definitive ‘queuing discipline’ protocol that will work for everyone as no two networks that are the same. Select and apply each queuing discipline in turn and test at dslreports.com/speedtest
After you have everything set run the speed test again. If you get near 90% of your previous measurement in each direction then things are cool. If you get results which are way off then chances are that you have reversed these values. You must enter a value for the uplink field but if you want you can enter 0 for the downlink field in which case no QoS will occur in that direction, setting your downlink field to 0 isn't recommended.
Now you have the router speeds set correctly, test for bufferbloat using the different ‘queuing discipline’ protocols available It probably bugs you to set less than 100% of your available bandwidth in these fields but this is required. There will be a bottleneck somewhere in the system and QoS can only work if the bottleneck is in your router where it has control. The goal is to force the bottleneck to be in your router as opposed to some random location out on the wire over which you have no control. Some ISP's even have bursting ("powerboost") which will temporarily give you extra bandwidth when you first start using your connection but will later throttle down to a sustained rate. Fortunately, there is usually a minimum level that you receive on a consistent basis and you must set your QoS limits below this minimum. The problem is finding this minimum, and you may have to repeat speed tests many times before determining it. For this reason start with 80% of your measured speed and try things for a couple of days. If the performance is acceptable you can start to inch your levels up. If you go even 2% higher than you should be, your QoS will totally stop working (just too high) or randomly stop working (when your ISP node/DSLAM is slow aka saturated). This can lead to a lot of confusion on your part so get it working first by conservatively setting these speeds and then optimise them later.
Our Basic Settings – Tested on our slowest line - TalkTalk Fibre 70mbps
Our Results BEFORE QoS Settings Implemented
Note the bufferbloat score of C which impacts the other scores and causes us many issues
Our Results AFTER QoS Settings Implemented
Note the slightly slower speeds but quality of connection improvement is massive. This equates to faster internet usage across our whole network
An in-depth guide to dd-wrt Quality Of Service can be found here: https://wiki.dd-wrt.com/wiki/index.php/Quality_of_Service
Conclusions
Broadband performance, seamless video streams and an uninterrupted service is not just about maximum speed. It is also about latency and delay. This often gets forgotten in the drive for greater and greater speeds. For browsing, online gaming, video streaming, voice, video telephony and conferencing, latency affects performance and the end-user experience far more than speed. Time spent 'tweaking' your router is time well spent.
i'd heard about this buffer thing before and now i understand it. made the changes to my Guardian Router as suggested and so far connection seems much more stable. Did take a little while testing different settings but well worth the effort.
I used the suggested settings and no longer get any buffering on my IPTV. Well worth the 5 min read on a very interesting arcticle. Thanks Guys