Sophist Member
  • Content count

  • Joined

  • Last visited

  • Speed Test

    My Results

Sean last won the day on January 11

Sean had the most liked content!

1 Follower

About Sean

  • Rank
    Sr. Member

Profile Information

  • Gender
  • Location

Recent Profile Visitors

4,702 profile views
  1. Lately when I have been trying any multi-threaded test, it has been quite erratic. It generally jumps straight to 25-50% the instant I start the test and it quite often gets stuck around the 80-90% mark. This is based on picking the UK option and testing on various devices (phone, laptop, desktop, etc.) and connections (home, work, LTE, etc.). As the test in progress screen shows the list of hosts it tests with, I took a screenshot of it and pinged all the hosts it showed. The culprit appears to be the UK2 tests: The hosts starting with 'uk.', i.e.,,, and all pinged with an average of 23ms, so these seem to be fine. The and both pinged with an average of 141ms, so these seem to be pointing to US or other long distance hosts. The, and hosts all fail to resolve the host name. This probably explains why the test skips the first 25-50% as the script probably detects the failed host lookups as these pieces completing their download. The timed out, even when I tried this in my web browser. This probably explains why it gets stuck around the 80-90% mark. Update: An example from Fiddler showing the host issues while a multi-threaded test takes place:
  2. It looks like they are doing some sort of traffic shaping, similar to the Irish Three network. First try using a different DNS such as Google ( and to rule out a slow DNS. The next suggestion would be to try the Data Saver plug-in with the PC's Chrome browser. This will route your browsing traffic through Google's server, which is likely to be prioritised. Unfortunately this will not improve downloads, which the browser does directly. Another workaround is try a VPN service that uses OpenVPN. As OpenVPN traffic is carried over port 443 (same as HTTPS), most ISPs will give the it the same priority as HTTPS. If secure websites such as Amazon load up pretty quick, there is a good chance an OpenVPN based connection will get around the throttling. HMA is one I know that supports connecting over pretty much any port number with OpenVPN. So if the ISP is prioritising port 8080 (to prioritise Ookla Speedtest traffic), just make the VPN connection over port 8080. I used this trick last year with the Three network until it stopped prioritising certain port numbers.
  3. In Ireland, the Three mobile network appears to have recently started prioritising YouTube traffic. Interestingly it no longer prioritises port 443 (SSL) or 8080 (Ookla's speedtest) like previously, so there's now a substantial difference between the speed I get on YouTube and even Ookla's Speedtest with Three. As Google Photos seems to share some resources with YouTube for video uploads, it also delivers unusually quick transfers from Google Photos during peak time. The following shows an example where I ran a speed test on Ookla (using a Dublin server) and started downloading a video from Google Photos: The following is the Ookla Speedtest result from the above screenshot and TestMy: Multi-threaded connections have always performed better on the Three network, so Ookla's figure tends to be several times higher during peak time than what can be achieved with simple web downloads and FTP transfers as how TestMy measures. The following shows another example, this time a file download from Leaseweb's test file site: 133KB/s x 8 = 1064Kbps, similar to the above TestMy download result. Now let's download a video clip I have in Google photos: 4MB/s x 8 = 32Mbps! It shows just how much bandwidth a provider can squeeze in during congestion. I have observed a similar variation on iD Mobile, a virtual mobile operator that operates on the Three network.
  4. Here in Ireland, it's the same issue. The major provider Three considers anything over 1Mbps as a non-issue with its 4G LTE broadband service and I've known people who had a tough time getting out of a lengthy contact. To make matters worse, they throttle per connection. As is multi-threaded, it will produce substantially higher test results than what one realistically gets, such as with YouTube. I also have a hard time trying to explain this to even supposedly technical people who just tell me that TestMy is not measuring properly as it's based in London and Speedtest uses servers in Ireland. The following shows an example one evening with a Three LTE connection: When I tried playing a handful of YouTube videos with the connection, it only spiked above 1Mbps a few times in the 'Stats for nerds': As ridiculous as it sounds, I am changing to a provider that only serves 3G in my area, to supplement my slow 4Mbps DSL connection. Unlike most providers, Meteor charges based on usage - €30 per 50GB prepaid. Although this works out more expensive than Three for heavy users, the difference is like night vs day in the evening: For anyone curious, 3G has no problem playing 1080p @ 60FPS over 3G in the evening my area.
  5. My 4G (LTE) based Internet connection is showing a clear example of this variation this evening. What's very clear when this happens is that the browser download speed (single connection) is a lot slower than what the multi-threaded tests (e.g. show. First tests - TestMy linear and multi-threaded: - Going by the Windows Resource Monitor, it made four simultaneous connections during the speed test. Google Fiber speed test - Going by the Windows Resource Monitor, it made 20 simultaneous connections during the test. Windows Task Manager Ethernet graph while downloading a Linux ISO file from Heanet (Ireland) - I know for certain the bottleneck is not on their end.
  6. The first thing I suggest is try the router in a few locations about the house to find where it gets the best signal reading. If your windows or glass are fairly new, the glass is likely Low-E coated, which radio signals cannot pass through. So it's worth checking other locations if you only tried on window sills. 4G (LTE) operates using a combination of vertical and horizontal polarisation (known as MIMO) and in turn requires two antennas to pick up both polarisations, i.e. one antenna aimed vertically and the second antenna aimed horizontally. What you can try next is aim one TP-Link antenna straight up and the other one sideways and try a download speed test. Then turn the router by 30 degrees and run another test and once more another 30 degrees. Although the signal strength figure may not change turning the router, running speed tests at a few angles will help find out which way the horizontal antenna needs to face as it is directional, similar to when an FM radio antenna is positioned horizontally. An external antenna needs to be placed in the loft or outside to provide much of a benefit, so I would suggest seeing what you can first get with its own antennas. Unless you plan going for a plan with a large download allowance (50GB+), I wouldn't worry too much about the speed. With 30Mbps, you can download 1GB in just under 5 minutes. If you manage to get 60Mbps, you can wipe out a 50GB allowance in just under 2 hours of non-stop downloading. The maximum speed you can get while the network is idle (off-peak) will depend on the bandwidth the mast is operating at. Based on your DSL speed, I assume you are in a fairly rural location, so you are likely picking up the 800MHz band, which usually operates on 10MHz bandwidth. Based on my experience with the Irish networks, this band peaks just over 50Mbps. From what I heard, the higher bands (1800MHz+) can peak about 140Mbps when operated on LTE Category 4 with 20MHz of bandwidth. Higher LTE categories (e.g. Cat 6) use band aggregation and requires a suitable router to take advantage of the bandwidth (e.g. Huawei E5186), but this is mainly only used in built-up high population areas due to the short range the higher frequency signals reach.
  7. This reminds me of something I must try some day. At my workplace there is an old D-Link router that failed where it kept randomly dropping connections and had high packet loss - It even struggled bringing up its web interface pages in the end. One thing I remember at the time was that gave the impression that the router was working fine consistently delivering the same speeds near what the DSL line was synced at. Yet, if I tried streaming YouTube, it would constantly get stuck buffering. I'm fairly sure it used to similarly get stuck on downloads. So if I get some spare time at work (e.g. lunch break when it's quiet), I'm curious to try running tests with the router and do a screen recording. I had several D-Link routers fail in a similar way in the past where they would randomly start dropping connections or cause Wi-Fi drop-outs, usually with high packet loss in extended ping tests.
  8. To me it seems like either a device/driver issue or a system process acting up. If this issue just started in the past day or two, did you install any software or hardware recently, e.g. new antivirus product, new printer, etc.? If you have, try uninstalling the software to see if the problem clears itself. If you didn't install or upgrade anything recently, another possibility is that a Windows/software update come in that started this issue such as a device driver update, in which case I would suggest trying a process of elimination by temporarily disabling the Windows start-up processes as follows: 1. Press the Windows Key + 'R', type in "msconfig" and click "OK". 2. Tick 'Selective startup' and go into the 'Services' tab. 3. Tick the lower box 'Hide all Microsoft services'. 4. Click 'Disable all', then click 'OK' and reboot the PC. 5. Check whether the problem has cleared at this point. If it hasn't, undo this change by going back into msconfig (Step 1) and tick 'Normal startup', i.e. skip the rest of this procedure.. 6. If the pause/delay has cleared, go back into msconfig (step 1), then into the 'Services' tab and tick half the boxes, click 'OK' and reboot the PC. 7. If the problem has not returned, repeat the above step again ticking half the remaining boxes, reboot and check again and so on until the problem occurs again. 8. When the problem occurs, start by unchecking the last checked box one at a time, rebooting until checking until the problem stops occurring. 9. Finally once you find the culprit service, check all the boxes except for the culprit service, so everything but the culprit service is running. The culprit service name (try Googling it if not clear) should indicate what is causing the issue. If the above procedure did not work, i.e. the problem was still present after step 5, then try unplugging everything but the keyboard, mouse and monitor to rule out any external device. Finally, if you have your Antivirus subscription details handy (if a paid product), try uninstalling your antivirus software temporarily to rule it out. I have seen a few past cases where McAfee caused strange symptoms, in one case giving the impression of a failed hard disk.
  9. I suggest getting a SIM for each of the prepay providers and run some speed tests during peak time, i.e. 8pm to 10pm. 4G can be surprisingly quick, particularly on cell towers with low contention and those operating on larger bandwidths such as 20MHz. My fastest speed test with TestMy on 4G was 102Mbps. On the other hand, if there are a lot of users in the area on the one provider, then the speed can severely plummet, particularly if if many are streaming at the same time. For example, I am currently using the Three 4G network to supplement my DSL connection. The speed test on the left is at the moment (off-peak) and the right speed test was yesterday evening near midnight (peak): For comparison, my DSL connection usually hits 3.9Mbps regardless of the time of day.
  10. From a quick check in TCPView, Google fiber's speed test is multi-threaded. Based on a test my with my connection (which only peaks about 45Mbps), it made 20 simultaneous connections: Although TestMy has a multi-threaded test option (default test is a single connection), the 200MB maximum block size is not sufficient to properly test Gigabit connections as the test will complete in under 2 seconds (900Mbps = 112.5MB per second). So for measuring your peak speed, the Google fibre test will likely be more accurate. On the other hand, the normal TestMy linear test will give an idea of what your connection is capable of with a single connection, similar to downloading a very large file with a web browser or FTP transfer. If you are getting in the 800Mbps range, your connection is fine. However, if it's much lower, e.g. below 500Mbps, then there is probably something limiting what you can achieve over a single connection, which the multi-threaded speed tests don't show, in which cause you would only be able to achieve the multi-threaded test result with a multi-threaded download manager, e.g. Firefox's DownThemAll plug-in which splits a large download into multiple segments and downloads these segments simultaneously.
  11. This issue seems to have resolved itself since posting as it now shows the TID graphs on all my devices.
  12. As far as I'm aware of, TestMy calculates its averages based on all test results for a given host, city, region, etc. Ookla's Speedtest awards are higher, but they only use the results from the fastest 10% of the test results: So peak time congestion dips, line issues, etc. will not impact an ISP's "average" performance as long as at least 10% of customers were getting what they are paying for. So a fixed wireless ISP could rank pretty well even if over half the customers get a tiny fraction of what they are paying for, as long as 10% of the speed tests were conducted off-peak or by customers on rural lightly loaded cell towers.
  13. When I click any of the TID links on the test results page, it shows a blank pop-up: It also does the same thing on my phone's browser. The test result graph shows when the test completes, this in case on my DSL line. However, when I go to the test results and touch the TID link, it shows up blank:
  14. Test run on the Irish Three 4G network in Donegal town, possibly LTE+ (OnePlus 2 phone): It's also my fastest TestMy download result to date, certainly did not expect to see my first >100Mbps result on a cellular network, let alone the Three network especially with the past experience of prioritising/throttling ports. As far as I can tell, they treat port 80 and 8080 equally now and the above test was a normal linear HTTP test with the UK server.