Welcome Guest ( Log In | Register )

Bump Topic Topic Closed RSS Feed
495 Pages « < 160 161 162 163 164 > » Bottom

Outline · [ Standard ] · Linear+

Unifi Official TM UniFi High Speed Broadband Thread V42, READ 1ST PAGE FOR RELEVANT WIFI INFO!

views
     
PRSXFENG
post Jun 17 2024, 04:19 PM

Look at all my stars!!
*******
Senior Member
2,608 posts

Joined: Nov 2020


QUOTE(NagaK @ Jun 17 2024, 02:48 PM)
Someone share shoppee link? I'm not fan of rack fans just another mess
*
Just search for "router fan" and you'll find plenty
enduser
post Jun 17 2024, 06:22 PM

Regular
******
Senior Member
1,280 posts

Joined: Jan 2003
QUOTE(Raymond T. @ Jun 17 2024, 12:54 PM)
Currently on Unifi 800mbps @ RM 230
*
Thar price for broadband only or with unifi tv ?

crackedbox
post Jun 17 2024, 06:49 PM

Getting Started
**
Junior Member
119 posts

Joined: Apr 2017
QUOTE(jiaen0509 @ Jun 16 2024, 06:23 PM)
I just bought this from Shopee with RM6.12 including shipping. So far use more than 2 weeks and the speed still constantly more than 1gbps.
user posted image
*
Hi, have one unit open box never used 2.5gb Unifi ONU unit for sale. Only one unit available. First come first serve. Do PM let me know if interested.
Raymond T.
post Jun 17 2024, 07:25 PM

Glock 19 Holder
******
Senior Member
1,368 posts

Joined: Aug 2010
QUOTE(enduser @ Jun 17 2024, 06:22 PM)
Thar price for broadband only or with unifi tv ?
*
With stupid Aneka which they say cannot unsubscribe
kwss
post Jun 17 2024, 07:47 PM

Regular
******
Senior Member
1,207 posts

Joined: Aug 2018
The slow link from South East Asia to US
I have tested this problem affect Malaysia and Singapore as a whole. If you have CDN in Malaysia / Singapore, but Origin server in the US, transfer will be very slow and almost certainly timeout.

This investigation started because I have a lot of problem working on Fedora, a flavor of Linux distro.

A little background:
There has been a lot of fights between Tier 1 and Tier 2 providers in EU and Asia lately, all of them involving Cogent. In the beginning of the year, Cogent de-peer NTT in EU, causing all EU traffic to peer in the US, increasing latency even if you are in EU. Example, two different AS in Amsterdam must peer in US, even when they are physically in the same city.
Last month, Cogent de-peer TATA in Asia, cutting off many single-homed customers from each other.
Submarine cable breakage towards EU and Japan also adds to the problem.

Example:
The website is fronted by Cloudflare or AWS CloudFront, and pings are fine. But when you actually starts transferring bulk data, speed quickly drop to less than 10KB/s and depending on how much data is being transferred, you will get a timeout. This happens regardless of the location of CDN either in KL, Johor or Singapore.

In Fedora's case, they are fronted by AWS CloudFront but their Origin is physically in the US.

Observation:
Anecdotes suggest the link will mostly be fine between 4am to 11am Malaysia Time. As you can see, I don't have time to wait for the stars to align.

My solution:
Since I have built my own anti-censorship "VPN", I use it to solve my problem. In AWS, all inter-AZ or inter-region traffic will be routed via what is known as "cold potato routing".
So I spin up one EC2 instance in Oregon and CloudFront it. I then connect to CloudFront KL, which will cold potato route me to my EC2 in Oregon.

The end result:
Hot potato routing Oregon -> TM
CODE

[ec2-user@i-063f1764f181fb8c3 ~]$ tracepath 115.134.191.254
1?: [LOCALHOST]                      pmtu 9001
1:  ip-172-23-48-1.us-west-2.compute.internal             0.083ms pmtu 1500
1:  no reply
2:  240.0.8.65                                            0.290ms
3:  242.0.65.69                                           4.915ms asymm  6
4:  240.1.228.15                                          9.178ms asymm  5
5:  242.4.195.71                                          6.556ms asymm 18
6:  150.222.214.74                                        6.814ms asymm 19
7:  as6939.10gigabitethernet9-9.core1.sea1.he.net         9.802ms asymm 24
8:  no reply
9:  as6939.100gigabitethernet2-3.core1.sjc2.he.net       23.471ms
10:  no reply
11:  telekom-malaysia-inc.e0-2.core3.sjc2.he.net         248.402ms asymm 17
12:  115.134.191.254                                     235.339ms reached
    Resume: pmtu 1500 hops 12 back 17


Hot potato routing TM -> Oregon
CODE

 tracepath ec2-18-246-47-189.us-west-2.compute.amazonaws.com
1?: [LOCALHOST]                      pmtu 1500
1:  _gateway                                              4.103ms
1:  _gateway                                              6.127ms
2:  _gateway                                              2.176ms pmtu 1480
2:  115.134.191.254                                      11.807ms
3:  10.55.49.97                                          41.665ms asymm  6
4:  10.55.100.228                                         9.979ms
5:  10.55.209.61                                         11.560ms
6:  ix-hge-0-0-0-22.ecore1.esin4-singapore.as6453.net    63.331ms asymm 16
7:  if-bundle-33-2.qcore1.esin4-singapore.as6453.net    276.803ms asymm 16
8:  if-bundle-19-2.qcore2.svw-singapore.as6453.net      195.879ms asymm 13
9:  no reply
10:  no reply
11:  if-et-24-2.hcore2.kv8-chiba.as6453.net              121.413ms asymm 12
12:  no reply
<----- snip ----->
30:  no reply
    Too many hops: pmtu 1480
    Resume: pmtu 1480


Cold potato routing EC2 Oregon -> CloudFront KL
CODE

[ec2-user@i-063f1764f181fb8c3 ~]$ tracepath 18.68.55.119
1?: [LOCALHOST]                      pmtu 9001
1:  ip-172-23-48-1.us-west-2.compute.internal             0.070ms pmtu 1500
1:  no reply
2:  240.0.8.65                                            0.321ms
3:  242.0.64.71                                           1.648ms asymm  6
4:  100.100.22.74                                         0.516ms asymm  6
5:  150.222.246.46                                      166.174ms asymm 22
6:  240.5.48.12                                         165.231ms asymm 22
7:  240.5.48.23                                         164.835ms asymm 22
8:  240.65.60.195                                       164.776ms asymm 22
9:  no reply
<----- snip ----->
30:  no reply
    Too many hops: pmtu 1500
    Resume: pmtu 1500


The ping! (Cold potato routing)
CODE

[ec2-user@i-063f1764f181fb8c3 ~]$ ping -c 10 18.68.55.119
PING 18.68.55.119 (18.68.55.119) 56(84) bytes of data.
64 bytes from 18.68.55.119: icmp_seq=1 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=2 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=3 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=4 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=5 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=6 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=7 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=8 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=9 ttl=234 time=165 ms
64 bytes from 18.68.55.119: icmp_seq=10 ttl=234 time=165 ms

--- 18.68.55.119 ping statistics ---
10 packets transmitted, 10 received, 0% packet loss, time 9010ms
rtt min/avg/max/mdev = 165.044/165.088/165.146/0.026 ms


The ping! (Hot potato routing)
CODE

[ec2-user@i-063f1764f181fb8c3 ~]$ ping -c 10 115.134.191.254
PING 115.134.191.254 (115.134.191.254) 56(84) bytes of data.
64 bytes from 115.134.191.254: icmp_seq=1 ttl=239 time=249 ms
64 bytes from 115.134.191.254: icmp_seq=4 ttl=239 time=263 ms
64 bytes from 115.134.191.254: icmp_seq=5 ttl=239 time=253 ms
64 bytes from 115.134.191.254: icmp_seq=6 ttl=239 time=229 ms
64 bytes from 115.134.191.254: icmp_seq=7 ttl=239 time=246 ms
64 bytes from 115.134.191.254: icmp_seq=8 ttl=239 time=262 ms
64 bytes from 115.134.191.254: icmp_seq=10 ttl=239 time=229 ms

--- 115.134.191.254 ping statistics ---
10 packets transmitted, 7 received, 30% packet loss, time 9159ms
rtt min/avg/max/mdev = 228.644/247.148/263.060/12.963 ms


Quantifiable improvement:
Packet loss reduce from 30% to 0%. Infinite improvement!
Latency reduce by 82ms, a 33% improvement.
Jitter reduce by 99.8%. Gamer rejoice!


This post has been edited by kwss: Jun 17 2024, 09:48 PM
jonathanwhm
post Jun 17 2024, 09:04 PM

Getting Started
**
Junior Member
87 posts

Joined: Jun 2011
QUOTE(NagaK @ Jun 17 2024, 03:48 PM)
Someone share shoppee link? I'm not fan of rack fans just another mess
*
https://shopee.com.my/YOHESO-12CM-Router-Mo...2015.7380199432


tng55
post Jun 17 2024, 09:04 PM

Regular
******
Senior Member
1,441 posts

Joined: Sep 2021


QUOTE(kwss @ Jun 17 2024, 01:08 PM)
Did you try setting MTU to 1280?
*
why need setting 1280 its not worth because ip address no change sad.gif
kwss
post Jun 17 2024, 09:05 PM

Regular
******
Senior Member
1,207 posts

Joined: Aug 2018
QUOTE(tng55 @ Jun 17 2024, 09:04 PM)
why need setting 1280 its not worth because ip address no change  sad.gif
*
I don't know if I want to explain to you or not. Nevermind, do what you like.
Ashren
post Jun 17 2024, 10:16 PM

Getting Started
**
Junior Member
156 posts

Joined: Dec 2010
QUOTE(tng55 @ Jun 17 2024, 09:04 PM)
why need setting 1280 its not worth because ip address no change  sad.gif
*
Forget about lowering your MTU just go for the highest supported by your isp ie TM on PPPoE is 1492. That's where fragmentation come into play. There's no need to go into that. Stick with 1492 or you will have packet issues with your connected devices.

As for the ip range nothing can be done unless TM sort their area gateway properly vs total subscribers in that area and distance to the nearest exchange. They cant even deploy a proper standard on the OLT's. Some areas get the broken Fiberhome OLT and some get the less problematic ZTE. Port sharing in crowded areas is also a problem that contribute to the issue.

You can tell now those with 42.xxx prefix are mostly either far south or up north. Very little to none in Klang Valley area nowadays. (Where are you from btw?)

Do a traceroute and take note on the first 2-4 local hops and compare it with two different ip range and two different area. Most of the time it won't be the same. Hence the variable in latency and performance to the same site/server between two ip range and location. It has got nothing to do with TM international routing or peering. All ip get the same international treatment. Its the local gateway.

Try this site and tell me what's your hop count under TCP/IP Fingerprint. Mine's 16.

https://browserleaks.com/ip

This post has been edited by Ashren: Jun 17 2024, 11:01 PM
Azims
post Jun 17 2024, 10:31 PM

Getting Started
**
Junior Member
211 posts

Joined: Oct 2014
QUOTE(Ashren @ Jun 17 2024, 10:16 PM)
Forget about lowering your MTU just go for the highest supported by your isp ie TM on PPPoE is 1492. That's where fragmentation come into play. There's no need to go into that. Stick with 1492 or you will have packet issues with your connected devices.

As for the ip range nothing can be done unless TM sort their area gateway properly vs total subscribers in that area and distance to the nearest exchange. They cant even deploy a proper standard on the OLT's. Some areas get the broken Fiberhome OLT and some get the less problematic ZTE. Port sharing in crowded areas is also a problem that contribute to the issue.

You can tell now those with 42.xxx prefix are mostly either far south or up north. Very little to none in Klang Valley area. Because you know why? Find out yourself. (Where are you from btw?)

Do a traceroute and take note on the first 2-4 local hops and compare it with two different ip range and two different area. Most of the time it won't be the same. Hence the variable in performance to the same site/server between ip range and location. It has got nothing to do with TM international routing or peering. All ip get the same treatment.

Try this site and tell me what's your hop count under TCP/IP Fingerprint. Mine's 16.

https://browserleaks.com/ip
*
will a public IP have this problem, or is it specific to CGNAT?
Ashren
post Jun 17 2024, 10:39 PM

Getting Started
**
Junior Member
156 posts

Joined: Dec 2010
QUOTE(Azims @ Jun 17 2024, 10:31 PM)
will a public IP have this problem, or is it specific to CGNAT?
*
The problem is on public ip but only to 42.xxx ip range. No other public ip range that Im aware of have this issue.

Those who got assigned to that ip range gateway somehow treated like behind a cgnat system. You cant get out from that ip range no matter how many times you reconnect to the gateway. Like stuck in some broken internal cgnat LAN. No other subscribers can get that ip range vice versa, its not in the public range rotation. Isnt it weird? I was on that ip range before covid hit for over a year and my ping to vultr sg playing call of duty always hover around 28-30ms. The very day I get a 175.xxx ip my ping goes down to 10-12ms until now. I'm speaking from experience I know what that fellow forumer feels like locked to that broken ip range forever for no apparent reason.

This post has been edited by Ashren: Jun 17 2024, 11:07 PM
TubeNRibbon
post Jun 17 2024, 11:11 PM

Getting Started
**
Junior Member
176 posts

Joined: May 2008
I added a fan on top of the white D-Link ONR. Still survive no speed drop for 2 days. I'm using bridge mode connected to my own router with 2.5G port.


jiaen0509
post Jun 17 2024, 11:32 PM

Look at all my stars!!
*******
Senior Member
3,305 posts

Joined: Dec 2012
QUOTE(TubeNRibbon @ Jun 17 2024, 11:11 PM)
I added a fan on top of the white D-Link ONR. Still survive no speed drop for 2 days. I'm using bridge mode connected to my own router with 2.5G port.
*
Let’s monitor again. Maybe later remove the fans and see whether speed drop occurs again or not.
excaliburcombo
post Jun 18 2024, 11:40 AM

Getting Started
**
Junior Member
222 posts

Joined: May 2016


Mine after 1Gbps Pro Upgrade. Last time 800Mbps YEP RM139(internet rm109+aneka rm30). Now RM 208.90 (internet rm198+ultimate max promo rm0.90+netflix premium upgrade rm10) which no price change(the extra RM10 discount goes to netflix premium upgrade) with previous package. How about the other member that complete upgrade to 1Gbps?

Attached Image
tng55
post Jun 18 2024, 12:18 PM

Regular
******
Senior Member
1,441 posts

Joined: Sep 2021


QUOTE(Ashren @ Jun 17 2024, 10:16 PM)
Forget about lowering your MTU just go for the highest supported by your isp ie TM on PPPoE is 1492. That's where fragmentation come into play. There's no need to go into that. Stick with 1492 or you will have packet issues with your connected devices.

As for the ip range nothing can be done unless TM sort their area gateway properly vs total subscribers in that area and distance to the nearest exchange. They cant even deploy a proper standard on the OLT's. Some areas get the broken Fiberhome OLT and some get the less problematic ZTE. Port sharing in crowded areas is also a problem that contribute to the issue.

You can tell now those with 42.xxx prefix are mostly either far south or up north. Very little to none in Klang Valley area nowadays. (Where are you from btw?)

Do a traceroute and take note on the first 2-4 local hops and compare it with two different ip range and two different area. Most of the time it won't be the same. Hence the variable in latency and performance to the same site/server between two ip range and location. It has got nothing to do with TM international routing or peering. All ip get the same international treatment. Its the local gateway.

Try this site and tell me what's your hop count under TCP/IP Fingerprint. Mine's 16.

https://browserleaks.com/ip
*
16 hops
tng55
post Jun 18 2024, 12:22 PM

Regular
******
Senior Member
1,441 posts

Joined: Sep 2021


QUOTE(Ashren @ Jun 17 2024, 10:39 PM)
The problem is on public ip but only to 42.xxx ip range. No other public ip range that Im aware of have this issue.

Those who got assigned to that ip range gateway somehow treated like behind a cgnat system. You cant get out from that ip range no matter how many times you reconnect to the gateway. Like stuck in some broken internal cgnat LAN. No other subscribers can get that ip range vice versa, its not in the public range rotation. Isnt it weird? I was on that ip range before covid hit for over a year and my ping to vultr sg playing call of duty always hover around 28-30ms. The very day I get a 175.xxx ip my ping goes down to 10-12ms until now. I'm speaking from experience I know what that fellow forumer feels like locked to that broken ip range forever for no apparent reason.
*
seee that prefect 175.xxx can get down 10-12ms daam you so lucky
how possible my office 175.xxx biz smooth same area my home to walk shop take 30sec near that DP diffrent place
my home DP BOX beside too DP TO MY HOME macam 20metter+++ that fiber very good but ip range bad
PJng
post Jun 18 2024, 06:05 PM

10k Club
********
All Stars
12,039 posts

Joined: Oct 2017


just share, nothing to discuss, 100mbps free upgrade 300mbps, up time 77days, wait for 100 days up time
TP link ax20, IP 115.135.x.x, yes router IP is 100.x.x.x
i just click above link, hops is 17
user posted image
kwss
post Jun 18 2024, 06:17 PM

Regular
******
Senior Member
1,207 posts

Joined: Aug 2018
I get 18 hop! Waiting for expert to teach me how to interpret this.
I give screenshot complete with full location and IP address to help with interpretation.

user posted image
jonathanwhm
post Jun 18 2024, 06:31 PM

Getting Started
**
Junior Member
87 posts

Joined: Jun 2011
QUOTE(kwss @ Jun 18 2024, 07:17 PM)
I get 18 hop! Waiting for expert to teach me how to interpret this.
I give screenshot complete with full location and IP address to help with interpretation.

user posted image
*
18 hops too, from KL.
user posted image
ycs
post Jun 18 2024, 09:22 PM

MEMBER
*******
Senior Member
4,231 posts

Joined: Jan 2003
From: Selangor



mine is 19 hops on 60.50.x.x

495 Pages « < 160 161 162 163 164 > » Top
Topic ClosedOptions
 

Change to:
| Lo-Fi Version
0.0217sec    0.34    6 queries    GZIP Disabled
Time is now: 8th December 2025 - 11:23 PM