AIS Fiber - anyone have recent experience?

[Pib]

This post servers two purposes...to give a few more more speed test results to London for UK folks interested in getting AIS and to show how two different speedtesters testing to the same city within minutes of each other can give widely different results primarily of the different ways they test.

Below you see two speed testers results to London done at approx 7pm...a non-VPN connection...just regular connection....speed seems to be picking up somewhat to London...a time of day thing I guess. One test by Testmy.net and one test my Speedtest.net.

Now these two speedtests run their tests differently. Testmy.net focuses on running a "single thread/stream" test unless you change it's settings. But Speedtest.net used "multiple threads/streams (2 to 4)" to run their test....Speedtest.net basically operates like a download manager in testing using multiple streams to their test sites which almost always results in "glorious" results. But hey, that is just fine if the site your are connecting to and the application you are using supports multiple streaming. So both testmy.net and speedtest.net are valid tests, although they use different testing methods.

Testmy.net to London Result using AIS 50/10

Speedtest.net to London Result using AIS 50/10

[Pib]

Whether you believe in Speedtest.net results or not, Speedtest.net rates AIS as the fastest Thailand Internet Service Provided (ISP).

See below speedtest.net link for more info...and be sure to read their Methodology for the testing because they are only basing their rating on the Top 10% speed test scores. With only the Top 10% (fastest) of test results being used that why the speeds below will probaby seem really high for the majority of internet users in Thailand unless they have fiber or cable/DOCSIS. Since AIS has entered the Thailand ISP market with only high speed fiber optics plans where other Thai ISPs have a variety of low and high speed ASDL/VDSL, cable, and fiber optics plans it's understandable AIS wins the speed race. But the award is still a good indicator of which ISP has the fastest speed if you happen to live in the right area.

http://www.speedtest.net/awards/th/isp

[nickbkk]

You, sir, are awesome. Thank you so much for taking all the time and effort to post such detailed and useful information to the forums. This is exactly the kind of info I had hoped I would get.

I got AIS Fibre 50/10 last Friday....it's great. Big, big improvement over my True DOCSIS/cable 15Mb/1.5Mb plan which I will be cancelling fairly soon.

[nickbkk]

Sure. No servers run from home but I backup terabytes of data to the cloud.

Can I ask the OP why upload speed is important ? Are you running a server ?

[muratremix]

Sure. No servers run from home but I backup terabytes of data to the cloud.

Can I ask the OP why upload speed is important ? Are you running a server ?

I thought it was running on your AIS line first, not in sweden.

AIS routing is not optimized like 3BB, but it seems bandwidth throughput is much better.

Some European destinations goes via USA in AIS. Before they were routing via (Thailand ->) Hong Kong (and then Singapore, which was kind of stupid.)

[rickudon]

Have AIS Fibre, was ok after we complained about speed and they changed connections. However, the past month or so i frequently spend minutes watching little circles go round every day, before it times out. To be honest, not much of an improvement on my old 3BB 10mb service. Off to UK in a day or 2 so cannot be bothered to complain until i return.

[Pib]

Done some quick Speedtest.net and Testmy.net speed tests today/Saturday from 2:30pm to 2:43pm with my AIS Fibre 50 down/10 up plan. You will see big differences between the two speedtesters for some locations. Part of that difference is Speedtest.net defaults to multithread/multistream testing (like how download managers achieve higher download speeds) where Testmy.net is not a multithread/multistream tester unless doing a special setup.

Test To Location Speedtest Results Testmy Results

Bangkok 50.4 / 12.1 No Bangkok Server

Singapore 51.4 / 12.0 46.7 / 10.9

Los Angeles 47.9 / 11.9 9.7 / 6.4

Frankfurt 43.2 / 10.9 10.2 / 6.0

London 41.7 / 12.0 2.4 / 2.3 (Seems speeds to UK a frequently slow when using Testmy.net)

[gandalf12]

These tests have no real validity. They are a best indication but can change at a moments notice. I don't think though that people realise that. To get a good comparison multiple connections using the same end points would need to be run simultaenously

[Pib]

And for those folks who may want download speed for just larger files, such as 100MB size files, below are my AIS Fibre 50/10 results for downloading 100MB sized files from this Softlayer Data Centers site at 4pm, plus or minus just a few minutes

San Jose US 25.7Mb

Singapore 44.7Mb

Frankfurt 34.7Mb

London 37.2Mb

[wpcoe]

@Pib: Did you ever change the DNS servers when using the PureVPN app? Every time I modify the DNS settings via Windows, the next time the app starts, it resets them to Automatic.

Have you had to change the DNS server settings from Automatic since you switched to AIS Fibre?

[Pib]

@Pib: Did you ever change the DNS servers when using the PureVPN app? Every time I modify the DNS settings via Windows, the next time the app starts, it resets them to Automatic.

Have you had to change the DNS server settings from Automatic since you switched to AIS Fibre?

I've never had that problem with AIS Fibre or TrueOnline when using the PureVPN app. Running the app does not change whatever DNS settings I may have made in Windows. And I ran some more tests using the App today where I had specific DNSs (Google and OpenDNS) setup in Windows and the running the app would not cause those Windows DNS setting to revert to auto.

From some googling, here's a couple of things you might want to check for:

1. Maybe the PureVPN Secure DNS setting is activated (by default it is not...maybe you changed it). From reading the PureVPN support pages below (one for New App...one for Old App) it seems it forces the app to use a Google DNS and OpenDNS which may force your Windows DNS back to Automatic. Or maybe the webpage is saying it forces it from Automatic to the Preferrred DNS of Google or OpenDNS. Take a read and see how you interpret it.

https://support.purevpn.com/how-to-switch-to-secure-dns-in-purevpn-new-windows-software-beta

https://support.purevpn.com/what-is-securedns-and-how-to-activate-it

2. I doubt this is the problem or you would have noticed it before stating to use PureVPN, but maybe you have a router settings enabled that forces use of the ISPs DNS or whatever DNS you might enter in the Router DNS setting. Below is a cut and paste from a guy having a similar problem to you where his Windows DNS settings would always get changed...turns out his ISP's router was forcing the changes to Windows.

Not sure how that answer got selected as the solution, as I didn't select it, but it turns out that my router was the culprit. It had a setting to let my isp manage my dns address. I just unchecked it, put in my own, and now it's fine again. Not sure how that option got ticked in the first place, though.

[Pib]

I think many of us would agree Sunday, especially Sunday evening and night to around midnight, is the slowest internet time of the week in Thailand...I know it has been for me over the years here in western Bangkok.

Did some Speedtest.net and Testmy.net speed tests from around 8:45pm to 9:00pm tonight/Sunday with my AIS Fibre 50 down / 10 up plan I've now had for about 9 days.

Keep in mind Speedtest.net is a multi-thread/stream tester (2 to 4 threads to its test sites) where Testmy.net is a single-thread/stream tester to its test sites unless you reconfig the test. Speedtest.net generally gives more glorious results to many locations due to its different testing method using multi-streams kinda like a download manager.

Take below test results with a grain of salt....your results may (will) vary depending on your ISP, speed plan, time of day, etc.

Speedtest.net Results

Speedtest-Bangkok to Bangkok (just to test my local connection speed...ensure it's 50down/10up)

Speedtest-Bangkok to Singapore

Speedtest-Bangkok to Los Angeles

Speedtest-Bangkok to Frankfurt

Speedtest-Bangkok to London

Testmy.net Results

TestMy-Bangkok to Singapore

TestMy-Bangkok to Los Angeles

TestMy-Bangkok to Frankfurt

TestMy-Bangkok to London

Edited July 10, 2016 by Pib

[zib]

You can check http://slim.keff.org/speedtest/

It runs at 05:00, 12:00, 20:00, 00:00 every day

It bypasses the transparent proxy AIS has for HTTP. So should be accurate.

[wpcoe]

@Pib: Did you ever change the DNS servers when using the PureVPN app? Every time I modify the DNS settings via Windows, the next time the app starts, it resets them to Automatic.

Have you had to change the DNS server settings from Automatic since you switched to AIS Fibre?

I've never had that problem with AIS Fibre or TrueOnline when using the PureVPN app. Running the app does not change whatever DNS settings I may have made in Windows. And I ran some more tests using the App today where I had specific DNSs (Google and OpenDNS) setup in Windows and the running the app would not cause those Windows DNS setting to revert to auto.

From some googling, here's a couple of things you might want to check for:

1. Maybe the PureVPN Secure DNS setting is activated (by default it is not...maybe you changed it). From reading the PureVPN support pages below (one for New App...one for Old App) it seems it forces the app to use a Google DNS and OpenDNS which may force your Windows DNS back to Automatic. Or maybe the webpage is saying it forces it from Automatic to the Preferrred DNS of Google or OpenDNS. Take a read and see how you interpret it.

https://support.purevpn.com/how-to-switch-to-secure-dns-in-purevpn-new-windows-software-beta

https://support.purevpn.com/what-is-securedns-and-how-to-activate-it

2. I doubt this is the problem or you would have noticed it before stating to use PureVPN, but maybe you have a router settings enabled that forces use of the ISPs DNS or whatever DNS you might enter in the Router DNS setting. Below is a cut and paste from a guy having a similar problem to you where his Windows DNS settings would always get changed...turns out his ISP's router was forcing the changes to Windows.

Not sure how that answer got selected as the solution, as I didn't select it, but it turns out that my router was the culprit. It had a setting to let my isp manage my dns address. I just unchecked it, put in my own, and now it's fine again. Not sure how that option got ticked in the first place, though.

Ah, here we are!

I got the DNS server settings to stick. I think I was trying to modify the Windows setting while the connection was active and the app was open. I put in the Google DNS server addresses with the connection not active and the app closed and since then the Google DNS servers have remained.

[muratremix]

You can check http://slim.keff.org/speedtest/

It runs at 05:00, 12:00, 20:00, 00:00 every day

It bypasses the transparent proxy AIS has for HTTP. So should be accurate.

I thought you use 50/20 home pro package with dynamic IPv4. They still put you through transparent proxy??

[Pib]

Ah, here we are!

I got the DNS server settings to stick. I think I was trying to modify the Windows setting while the connection was active and the app was open. I put in the Google DNS server addresses with the connection not active and the app closed and since then the Google DNS servers have remained.

Yeap, no-can-change settings while the connection is active...a message of some type will appear saying no-can-do. Been there, attempted that.

The way you described it before I was under the impression you had successfully made changes but they would not stick once closing and reopening the app.

Glad you got it fixed. I find 9 out of 10 times my problems are always "operator error."

[Pib]

You can check http://slim.keff.org/speedtest/

It runs at 05:00, 12:00, 20:00, 00:00 every day

It bypasses the transparent proxy AIS has for HTTP. So should be accurate.

I thought you use 50/20 home pro package with dynamic IPv4. They still put you through transparent proxy??

I know a transparent proxy can be hard to detect, but when using the WhatIsMyIP Proxy Check Tool for me on my AIS 50/10 Home plan, it runs the check and says:" Success![/size] No Proxy Detected!"[/size]

But since VPN can bypass an ISP's proxy and some of my VPN speeds to certain locations, like to the UK, have been faster than a non-VPN connection maybe they are using a transparent proxy since transparent proxies seem to be the norm now days.

Also, I don't think I would consider a Transparent Proxy a bad thing as it does help to speedup internet use. Sure it can also track/limit your internet usage in certain ways, but hey, that's the world we live in now days.

Edited July 11, 2016 by Pib

[zib]

You can check http://slim.keff.org/speedtest/

It runs at 05:00, 12:00, 20:00, 00:00 every day

It bypasses the transparent proxy AIS has for HTTP. So should be accurate.

I thought you use 50/20 home pro package with dynamic IPv4. They still put you through transparent proxy??

The transparent proxy is always there regardless of connection. That's how they block the blacklisted sites (porn etc)

[zib]

You can check http://slim.keff.org/speedtest/

It runs at 05:00, 12:00, 20:00, 00:00 every day

It bypasses the transparent proxy AIS has for HTTP. So should be accurate.

I thought you use 50/20 home pro package with dynamic IPv4. They still put you through transparent proxy??

I know a transparent proxy can be hard to detect, but when using the WhatIsMyIP Proxy Check Tool for me on my AIS 50/10 Home plan, it runs the check and says:" Success![/size] No Proxy Detected!"[/size]

But since VPN can bypass an ISP's proxy and some of my VPN speeds to certain locations, like to the UK, have been faster than a non-VPN connection maybe they are using a transparent proxy since transparent proxies seem to be the norm now days.

Also, I don't think I would consider a Transparent Proxy a bad thing as it does help to speedup internet use. Sure it can also track/limit your internet usage in certain ways, but hey, that's the world we live in now days.

That test doesn't work. You need to test with "tcptraceroute". Don't know if it's available for windows. What you do is pick a website that's not in thailand (dont pick google.com or anything on cdn which will most likely be in thailand anyway).

One good example I can give you is "www.sunet.se".

---

# tcptraceroute www.sunet.se

Selected device eth0, address 172.17.11.1, port 34082 for outgoing packets

Tracing the path to www.sunet.se (192.36.171.231) on TCP port 80 (http), 30 hops max

1 172.17.11.2 0.483 ms 0.262 ms 0.268 ms

2 10.207.240.82 8.711 ms 8.188 ms 8.123 ms

3 10.207.240.81 19.126 ms 19.349 ms 18.638 ms

4 49.228.1.110 19.323 ms 24.929 ms 19.432 ms

5 49-230-34-78.ais-idc.com (49.230.34.78) 19.928 ms 18.990 ms 19.222 ms

6 webc.sunet.se (192.36.171.231) [open] 19.325 ms 19.450 ms 19.975 ms

---

Here you can see it's grabbed by the AIS proxy. Seen by both the jumps and the low latency. To be extra sure you can specify to to the traceroute on port 443 (ssl) instead :

# tcptraceroute www.sunet.se 443

Selected device eth0, address 172.17.11.1, port 57091 for outgoing packets

Tracing the path to www.sunet.se (192.36.171.231) on TCP port 443 (https), 30 hops max

1 172.17.11.2 0.433 ms 0.259 ms 0.239 ms

2 10.207.240.82 8.247 ms 8.245 ms 8.309 ms

3 10.207.240.81 18.624 ms 18.767 ms 18.683 ms

4 49.228.1.110 19.185 ms 31.995 ms 19.395 ms

5 ae54-514.IIG-TLS1-PE01.ais-idc.com (49.228.1.118) 19.221 ms 19.319 ms 26.841 ms

6 xe-0-1-1-0.IIG-TLS1-PE01.ais-idc.com (49.231.44.101) 38.713 ms 30.073 ms 29.506 ms

7 49-231-44-169.sbn-idc.com (49.231.44.169) 43.075 ms 42.541 ms 43.570 ms

8 124.158.228.85 53.916 ms 53.805 ms 53.591 ms

9 * * *

10 * * *

11 * * *

12 * * *

13 dk-uni.nordu.net (109.105.97.126) 334.379 ms 333.678 ms 333.934 ms

14 * * *

15 m1tug.sunet.se (109.105.102.18) 337.511 ms 336.489 ms 336.029 ms

16 webc.sunet.se (192.36.171.231) 336.233 ms 336.047 ms 335.549 ms

17 webc.sunet.se (192.36.171.231) [open] 336.539 ms 336.375 ms 337.062 ms

---

[zib]

Here's an example from 3BB DSL, Site in Sweden.

# tcptraceroute 130.244.0.1 80
Selected device eth0, address 172.17.11.1, port 34877 for outgoing packets
Tracing the path to 130.244.0.1 on TCP port 80 (http), 30 hops max
1 172.17.11.2 0.420 ms 0.273 ms 0.262 ms
2 192.168.1.1 1.151 ms 1.016 ms 0.998 ms
3 10.121.83.101 19.568 ms
10.121.83.177 19.702 ms
10.121.83.101 23.091 ms
4 10.121.83.178 19.114 ms 19.113 ms 20.093 ms
5 mx-ll-110.164.14-141.static.3bb.co.th (110.164.14.141) 31.062 ms 35.201 ms 29.412 ms
6 mx-ll-110.164.14-140.static.3bb.co.th (110.164.14.140) 29.930 ms 30.469 ms 29.907 ms
7 mx-ll-110.164.1-167.static.3bb.co.th (110.164.1.167) 29.560 ms 29.975 ms 29.004 ms
8 mx-ll-110.164.1-177.static.3bb.co.th (110.164.1.177) 30.625 ms 29.542 ms 29.280 ms
9 130.244.0.1 [open] 29.997 ms 29.906 ms 30.294 ms

[nickbkk]

I don't understand the reference to Sweden or, in fact, that entire first line...

Sure. No servers run from home but I backup terabytes of data to the cloud.

Can I ask the OP why upload speed is important ? Are you running a server ?

I thought it was running on your AIS line first, not in sweden.

AIS routing is not optimized like 3BB, but it seems bandwidth throughput is much better.

Some European destinations goes via USA in AIS. Before they were routing via (Thailand ->) Hong Kong (and then Singapore, which was kind of stupid.)

[Pib]

You can check http://slim.keff.org/speedtest/

It runs at 05:00, 12:00, 20:00, 00:00 every day

It bypasses the transparent proxy AIS has for HTTP. So should be accurate.

I thought you use 50/20 home pro package with dynamic IPv4. They still put you through transparent proxy??

I know a transparent proxy can be hard to detect, but when using the WhatIsMyIP Proxy Check Tool for me on my AIS 50/10 Home plan, it runs the check and says:" Success![/size] No Proxy Detected!"[/size]

But since VPN can bypass an ISP's proxy and some of my VPN speeds to certain locations, like to the UK, have been faster than a non-VPN connection maybe they are using a transparent proxy since transparent proxies seem to be the norm now days.

Also, I don't think I would consider a Transparent Proxy a bad thing as it does help to speedup internet use. Sure it can also track/limit your internet usage in certain ways, but hey, that's the world we live in now days.

That test doesn't work. You need to test with "tcptraceroute". Don't know if it's available for windows. What you do is pick a website that's not in thailand (dont pick google.com or anything on cdn which will most likely be in thailand anyway).

One good example I can give you is "www.sunet.se".

---

# tcptraceroute www.sunet.se

Selected device eth0, address 172.17.11.1, port 34082 for outgoing packets

Tracing the path to www.sunet.se (192.36.171.231) on TCP port 80 (http), 30 hops max

1 172.17.11.2 0.483 ms 0.262 ms 0.268 ms

2 10.207.240.82 8.711 ms 8.188 ms 8.123 ms

3 10.207.240.81 19.126 ms 19.349 ms 18.638 ms

4 49.228.1.110 19.323 ms 24.929 ms 19.432 ms

5 49-230-34-78.ais-idc.com (49.230.34.78) 19.928 ms 18.990 ms 19.222 ms

6 webc.sunet.se (192.36.171.231) [open] 19.325 ms 19.450 ms 19.975 ms

---

Here you can see it's grabbed by the AIS proxy. Seen by both the jumps and the low latency. To be extra sure you can specify to to the traceroute on port 443 (ssl) instead :

# tcptraceroute www.sunet.se 443

Selected device eth0, address 172.17.11.1, port 57091 for outgoing packets

Tracing the path to www.sunet.se (192.36.171.231) on TCP port 443 (https), 30 hops max

1 172.17.11.2 0.433 ms 0.259 ms 0.239 ms

2 10.207.240.82 8.247 ms 8.245 ms 8.309 ms

3 10.207.240.81 18.624 ms 18.767 ms 18.683 ms

4 49.228.1.110 19.185 ms 31.995 ms 19.395 ms

5 ae54-514.IIG-TLS1-PE01.ais-idc.com (49.228.1.118) 19.221 ms 19.319 ms 26.841 ms

6 xe-0-1-1-0.IIG-TLS1-PE01.ais-idc.com (49.231.44.101) 38.713 ms 30.073 ms 29.506 ms

7 49-231-44-169.sbn-idc.com (49.231.44.169) 43.075 ms 42.541 ms 43.570 ms

8 124.158.228.85 53.916 ms 53.805 ms 53.591 ms

9 * * *

10 * * *

11 * * *

12 * * *

13 dk-uni.nordu.net (109.105.97.126) 334.379 ms 333.678 ms 333.934 ms

14 * * *

15 m1tug.sunet.se (109.105.102.18) 337.511 ms 336.489 ms 336.029 ms

16 webc.sunet.se (192.36.171.231) 336.233 ms 336.047 ms 335.549 ms

17 webc.sunet.se (192.36.171.231) [open] 336.539 ms 336.375 ms 337.062 ms

---

How do you know it a proxy? Sure I see some hops with AIS in the name, but what they are doing I couldn't say. One has a name indicating it's a gateway. But heck, my ISP is now AIS, I'm on their network, so I would expect to be going through some of their servers/gateways/etc as they route it out of Thailand.

[zib]

How do you know it a proxy? Sure I see some hops with AIS in the name, but what they are doing I couldn't say. One has a name indicating it's a gateway. But heck, my ISP is now AIS, I'm on their network, so I would expect to be going through some of their servers/gateways/etc as they route it out of Thailand.

I can't explain it more than I did above. What you can see is that they redirect port 80 to their own servers (showed by the amount of jumps and latency, 19 ms to Sweden is impossible). The second traceroute for port 443 shows that this is not redirected.

[Pib]

zib,

Since you are on the AIS 50/20 "Pro" package which supposedly gives faster "international" speed over a Home program like my AIS 50/10 Home package, based on all the speed tests I've posted and you have also run but in a somewhat different way apparently using Linux vs Windows like I use, do you think the Pro package provides any additional speed over the Home package?

[zib]

zib,

Since you are on the AIS 50/20 "Pro" package which supposedly gives faster "international" speed over a Home program like my AIS 50/10 Home package, based on all the speed tests I've posted and you have also run but in a somewhat different way apparently using Linux vs Windows like I use, do you think the Pro package provides any additional speed over the Home package?

It might be just a trick. I haven't been able to confirm that PowerPro is any better internationally. But it might be that they cap the PowerHome after a while or that the PowerPro only gets priority (through qos) at peak hours. But currently I've seen nothing that proves the PowerPro is better. I got it because of the public ipv4.

I will order the PowerHome for my company (Home heh..funny) in the end of this month. It's not far from my house so I think I will be able to do fairly accurate international tests to compare them.

[zib]

Have AIS Fibre, was ok after we complained about speed and they changed connections. However, the past month or so i frequently spend minutes watching little circles go round every day, before it times out. To be honest, not much of an improvement on my old 3BB 10mb service. Off to UK in a day or 2 so cannot be bothered to complain until i return.

They did have a problem with their transparent proxy(s) during almost all of June. The problem was some MTU problem so sites that have PMTUD blocked wouldn't load.

The problem exists because of the AIS Fiber connections are using PPPOE which brings the MTU down to 1492 instead of the standard 1500. So without PMTUD websites tries to send packets that are too big so only parts of the data goes through and then the connection just hangs.

But seems to have been fixed in the end of June. Can't see the problems now when I test.

[Pib]

zib,

Since you are on the AIS 50/20 "Pro" package which supposedly gives faster "international" speed over a Home program like my AIS 50/10 Home package, based on all the speed tests I've posted and you have also run but in a somewhat different way apparently using Linux vs Windows like I use, do you think the Pro package provides any additional speed over the Home package?

It might be just a trick. I haven't been able to confirm that PowerPro is any better internationally. But it might be that they cap the PowerHome after a while or that the PowerPro only gets priority (through qos) at peak hours. But currently I've seen nothing that proves the PowerPro is better. I got it because of the public ipv4.

I will order the PowerHome for my company (Home heh..funny) in the end of this month. It's not far from my house so I think I will be able to do fairly accurate international tests to compare them.

A week or so ago when I talked to AIS Tech Support about some issues...really just curiosity questions about their Playbox and a couple about their internet, I asked them does the Pro package give faster speed to international sites. The tech rep just said "more stable" connections whatever more stable means. I asked him again does that mean more speed and he just answered again with the same more stable answer. I then moved on to my next curiosity question.

[zib]

zib,

Since you are on the AIS 50/20 "Pro" package which supposedly gives faster "international" speed over a Home program like my AIS 50/10 Home package, based on all the speed tests I've posted and you have also run but in a somewhat different way apparently using Linux vs Windows like I use, do you think the Pro package provides any additional speed over the Home package?

It might be just a trick. I haven't been able to confirm that PowerPro is any better internationally. But it might be that they cap the PowerHome after a while or that the PowerPro only gets priority (through qos) at peak hours. But currently I've seen nothing that proves the PowerPro is better. I got it because of the public ipv4.

I will order the PowerHome for my company (Home heh..funny) in the end of this month. It's not far from my house so I think I will be able to do fairly accurate international tests to compare them.

A week or so ago when I talked to AIS Tech Support about some issues...really just curiosity questions about their Playbox and a couple about their internet, I asked them does the Pro package give faster speed to international sites. The tech rep just said "more stable" connections whatever more stable means. I asked him again does that mean more speed and he just answered again with the same more stable answer. I then moved on to my next curiosity question.

That could imply priorityzation. So during non-peak hours both get the same speed. But during peak (18-22 probably) the PowerPro gets priority over the PowerHome connections. So more "stable" == same speed during all hours.

Since they are fairly new and just recently started pushing AIS Fiber more they probably have enough capacity now so Home or Pro Speed is not so noticable. If the customer base increases and they do same bandwidth management as awful for example True and 3BB (3bb worse) then we will probably notice it later

My company currently has 3BB and my house before had True. 3BB was unusable internationally during peak. True was better but varied alot day to day.

[Pib]

Here's what their website says about the differences.

[muratremix]

Here's an example from 3BB DSL, Site in Sweden.

# tcptraceroute 130.244.0.1 80

Selected device eth0, address 172.17.11.1, port 34877 for outgoing packets

Tracing the path to 130.244.0.1 on TCP port 80 (http), 30 hops max

1 172.17.11.2 0.420 ms 0.273 ms 0.262 ms

2 192.168.1.1 1.151 ms 1.016 ms 0.998 ms

3 10.121.83.101 19.568 ms

10.121.83.177 19.702 ms

10.121.83.101 23.091 ms

4 10.121.83.178 19.114 ms 19.113 ms 20.093 ms

5 mx-ll-110.164.14-141.static.3bb.co.th (110.164.14.141) 31.062 ms 35.201 ms 29.412 ms

6 mx-ll-110.164.14-140.static.3bb.co.th (110.164.14.140) 29.930 ms 30.469 ms 29.907 ms

7 mx-ll-110.164.1-167.static.3bb.co.th (110.164.1.167) 29.560 ms 29.975 ms 29.004 ms

8 mx-ll-110.164.1-177.static.3bb.co.th (110.164.1.177) 30.625 ms 29.542 ms 29.280 ms

9 130.244.0.1 [open] 29.997 ms 29.906 ms 30.294 ms

Interesting.

My 50/10 vdsl 3bb has no transparent proxy (or probably have??)

RT-AC56R (0.0.0.0) Mon Jul 11 13:13:45 2016

Packets Pings

Host Loss% Snt Last Avg Best Wrst StDev

1. mx-ll-110.164.14-215.static.3bb.co.th 0.0% 58 21.62.82.9 19.23.24.5 0.0

2. mx-ll-110.164.14-214.static.3bb.co.th 0.0% 5 19.1 29.75.68.80.53.9 16.7

3. mx-ll-110.164.1-167.static.3bb.co.th 0.0% 53 19.4 19.14 18.8 19.3 0.0

4. mx-ll-110.164.1-177.static.3bb.co.th 0.0% 41 22.26 25.3 19.18.81.5 0.0

5. 130.244.0.1 0.0% 4 228.4 209.07.30.72628.9 29.5

I used mtr --tcp -P 80 (HOSTNAME) command because I couldn't find tcptraceroute for mac or ipkg/opkg on NAS/Router.

without --tcp parameter, it is like normal traceroute;

Host Loss% Snt Last Avg Best Wrst StDev

1. mx-ll-110.164.14-215.static.3bb.co.th 0.0% 43 19.6 19.3 19.7 19.7 0.0

2. mx-ll-110.164.14-214.static.3bb.co.th 0.0% 4 19.6 19.0 18.6 19.6 0.0

3. mx-ll-110.164.1-167.static.3bb.co.th 0.0% 4 19.3 23.7 9.7.41438.2 0.0

4. mx-ll-110.164.1-123.static.3bb.co.th 0.0% 4 19.1 19.1 18.7 19.4 0.0

5. 192.168.255.250 0.0% 4 221.1 255.159.2.186.4.3 0.0

0.0% 4 221.8 223.73.81.74229.6 5.5

7. 217.239.42.81 0.0% 4 219.55219.4 219.3 219.4 0.0

8. 80.150.168.106 0.0% 4 225.8 229.4 225.81.98.5 0.0

9. ams-core-2.bundle-ether8.tele2.net 0.0% 4 230.4 230.5 230.4 230.8 0.0

10. gbg-core-1.bundle-ether1.tele2.net 0.0% 4 257.0 257.0 257.0 257.1 0.0

11. hgd-core-1.hundredgige0-7-0-1.tele2.net 0.0% 4 241.0 240.76240.4 241.0 0.0

12. 130.244.0.1 0.0% 3 243.3 243.4 243.3 243.4 0.0

No matter which destination I test, I get 220 ms on last hop. Even for singapore.

What kind of transparent proxy is this? True's transparent proxy was obvious. This one is insidious!

Edited July 11, 2016 by muratremix