True Maxnet Indy Vs Premier And Interenational Access

[Upcountry]

As I understand it, the difference between the Indy and Premier packages, besides the advertised speed increase for Premier is this:

1. Fewer users per "connection" (forgot the correct term) I.E., 20 per connection for Indy, and 10 per connection for Premier, which means less competition for bandwidth.

2. Better international bandwidth and less or no bittorrent/P2P throttling.

I upgraded to Premier last month, and for a while I did see a noticeable, though not dramatic, improvement. But for almost a week now it has been very bad.

Traceroutes are in the 3 digits, especially starting at the singapore.teleglobe.net hop. The response remains at 200 to 500 ms for the remainder of the route. Using a US based shell account I that I can telnet to, I did a traceroute back here to my IP and various Maxnet IP addresses. Response times were much faster, at least until the Singapore hop.

Something is strange about that*. Given the kind of results I'm getting right now, I'm thinking of going back to Indy, because seems like their statement about international access is a lie, or their policy may have changed along with the speed increases and whatever Big Brother activities that True may be engaged in.

Has anyone else noticed this?

*I should add that one reason for the slow access from here could be True's available bandwidth in general, but it was better last week.

Edited June 18, 2008 by Upcountry

[monty]

Can you post a trace over here?

Maxnet has it's own international gateway, but they also buy a lot of international bandwidth from CAT.

If there are CAT servers in the beginning of the trace, then it's not really Maxnet's fault, but CAT doing some creative routing!

If it's all Maxnet servers (registered under TT&T or triple T if you do a whois on the IP's), then it's their IIG doing messed up routing.

I remember several months ago TT&T trying to use their own gateway and it was a mess. Pings of over 1000 msec to the US along with loads of dropped packets. Within 24 hours they were back routing through CAT's gateway and everything was good again.

Can't see what they're up to now, since I'm on CAT ADSL now...

[Upcountry]

Thanks, Monty,

Lots of tttmaxnet.com hops. It does look like they are trying to get by with their own gateway again.

Also, check out the spooky 10.121.65.250 hop. It's not usually that long a delay.

Tracing route to europa.eu [147.67.119.102]

over a maximum of 30 hops:

1 <1 ms 1 ms 1 ms 192.168.1.1

2 126 ms 128 ms 66 ms mx-ll-58.147.40-1.tttmaxnet.com [58.147.40.1]

3 * * * Request timed out.

4 * * 2077 ms 10.121.65.250

5 24 ms 34 ms 16 ms mx-ll-58.147.0-25.tttmaxnet.com [58.147.0.25]

6 161 ms 65 ms 17 ms mx-ll-58.147.0-86.tttmaxnet.com [58.147.0.86]

7 18 ms 30 ms * mx-ll-58.147.0-54.tttmaxnet.com [58.147.0.54]

8 * * * Request timed out.

9 19 ms 23 ms 17 ms 218.100.47.20

10 129 ms 84 ms 68 ms t3-0-1-0.0.core2.seo2.flagtel.com [62.216.144.225]

11 * * * Request timed out.

12 439 ms 444 ms 504 ms so-0-1-2.0.pjr01.mmb004.flagtel.com [85.95.25.97]

13 * 176 ms 256 ms 85.95.25.77

14 357 ms 446 ms 287 ms ge-0-0-0.0.cjr01.alx001.flagtel.com [62.216.128.157]

15 409 ms 271 ms 282 ms t2a1.uk-lon1.eu.bt.net [195.66.224.108]

16 * 333 ms 389 ms t2c1-ge2-0-0.uk-lon1.eu.bt.net [166.49.135.101]

17 * * 256 ms t2c2-ge7-1.uk-lon2.eu.bt.net [166.49.208.5]

18 276 ms 389 ms * t2c1-p9-1.be-bru.eu.bt.net [166.49.164.54]

19 * * * Request timed out.

20 420 ms 308 ms 266 ms 166-49-157-158.eu.bt.net [166.49.157.158]

21 370 ms 270 ms 446 ms 212.8.176.130

22 * * * Request timed out.

23 * * * Request timed out.

... etc.

[Jimjim]

So what's the best superfast internet provider in Thailand according to all the experts?

[youngkiwi]

ksc.net

[monty]

218.100.47.20 is the actual Triple T international gateway!

Yep, they're not using CAT, but ping times are acceptable.

Yeah, these private IP range addresses showing is poor management. You normally hide them. It's been mooted that those are actually servers snooping up on your e-mails and surfing behavior, hence the long delays

I'd reckon CAT is still offering the best ADSL, although not available everywhere.

They do have the most experience (doesn't make them immune from the occasional mess up!), have loads of bandwidth and it makes sense they give routing priority to their own subscribers.

Most other ISP's source the majority of their bandwidth through CAT, so there's quite a few extra hops involved introducing inevitable delays.

And most other ISP's having their own gateway seem to be making an even bigger mess out of it then CAT is doing

Pings to US servers are usually just a tad over 200 msec, Thaivisa (Singapore located) comes in a between 60 and 70 msec...

[Upcountry]

Just got a call from a TTT tech (follow up from call I made four days ago). He asked for IP addresses that I have used for traceroute tests. He thought most hops were normal, but a few were higher. He also said the average was okay.

Monty, you thought the numbers were normal as well, but I swear it was much better last week. Perhaps you mean normal as in "snafu".

Some of my software client connections fail, and last night a internet-connected screen saver on my mac froze up the computer thanks to the delays.

I asked about the gateway. He confirmed that they are using their own gateway now, not CAT. I explained that I did a test by tracerouting from a shell account in the U.S. to Maxnet, and found that the results were much faster from there to here, than from here to there (I suppose that would be expected). He said that they plan to improve their gateway to Europe and the U.S. in the next couple of months or so. He also insisted that my connection and bandwidth is still better than Indy!

No confirmation, of course, about snooping on emails. He said they only cache web pages on port 80.

UC

[LivinLOS]

I have the premier cheapest (1050 per month.. Think it is listed as 1.5mbit but not sure, just had it installed 2 months ago)..

And its by far the best internet I have had in Thailand even faster than the CAT ADSL package. I am regularly pulling 1.8 / 1.9 Mbit to London and LA.. Almost 2 Mbits for full international bandwidth for 1000 baht per month..

[monty]

Well, it seems that the problems are related to their gateway, and not to you being on Premier or not.

Your ping times do look normal apart from that 10.xx.xx.xx range one.

Latency gets created by two things, physical distance (even at the speed of light it takes a little while to get round the earth) and delays at all the routers along the way.

Best case scenario with no delays at all the hops (there's actually delay at every hop, due to the processing speed of the equipment, with no delays I mean delays caused by overloaded equipment) to the US is about 200 msec.

It used to be that almost all Europe bound traffic got routed over the US, creating very long distances (it's a bit of a detour ) resulting in pings to Europe seldom being under 400 msec.

Now it seems that most Europe bound traffic goes from one of the Asian gateway countries (Singapore, Hong Kong, Japan) straight to Europe.

Probably your current problem are the companies TT&T's gateway buys their bandwidth from. Suddenly they don't have to buy from CAT anymore, so they tend to go for the cheapest suppliers which are often not the best! Once your data is in Singapore it is out of the hands of TT&T, and when you get big delays and dropped packets over there, it means that the supplier can't keep up with demand.

Singapore is roughly a 50 msec hop away, so anything longer within Singapore means overloaded servers. When servers are overloaded, your data ends up in a cue waiting to get processed (or even dropped altogether). Depending on how much TT&T pays them, will determine their QOS settings, which then determines priority of the TT&T originating traffic...

[Upcountry]

Monty, good description of routing issues, thanks.

Your ping times do look normal apart from that 10.xx.xx.xx range one.

Latency gets created by two things, physical distance (even at the speed of light it takes a little while to get round the earth) and delays at all the routers along the way.

.... I think you meant "strange" one, eh?

Three attempts, below, today. Note 4202 ms in one case. This is ridiculous and unacceptable, since it is under the control of the True (happening in the middle of all the other creepy tttmaxnet.com hops). It really must be the nefarious one. Maybe it is time we all started using encrypted email, etc., we value our privacy.

1 <1 ms <1 ms <1 ms 192.168.1.1

2 13 ms 12 ms 13 ms mx-ll-58.147.40-1.tttmaxnet.com [58.147.40.1]

3 * * * Request timed out.

4 * * 1990 ms 10.121.65.238

1 <1 ms <1 ms <1 ms 192.168.1.1

2 13 ms 12 ms 12 ms mx-ll-58.147.40-1.tttmaxnet.com [58.147.40.1]

3 * * * Request timed out.

4 2219 ms 4209 ms * 10.121.65.238

1 <1 ms <1 ms <1 ms 192.168.1.1

2 130 ms 31 ms 12 ms mx-ll-58.147.40-1.tttmaxnet.com [58.147.40.1]

3 * * * Request timed out.

4 * 2026 ms * 10.121.65.238

5 * 43 ms 163 ms mx-ll-58.147.0-25.tttmaxnet.com [58.147.0.25]