Telkom Internet Capped/Uncapped User Feedback (Pt2)

ponder

ponder

Honorary Master
Joined
Jan 22, 2005
Messages
92,823
1.PNG

http://www.speedtest.net/ actually fails with a "UPLOAD TEST ERROR. A socket error occurred during the Upload test. Please try again later." And that is on the durban saix speedtest.

Welcome to dial-up land!
 
Last edited:
B

Big Bean

Senior Member
Joined
Dec 4, 2009
Messages
583
streaker said:
That is strange indeed. Can you send me a message with your username and the IP's of the gaming servers ?
Click to expand...

During this period of issues, Dota 2 has not been affected but Counter Strike Global Offensive has had really high pings.
 
M

MickZA

Executive Member
Joined
Jan 19, 2007
Messages
7,575
Seems OK here in PE:

Ping: 34ms
Download: 9.19Mbps
Upload: 0.79Mbps

Code: 
[mick@house ~]$ ping -qc 20 bbc.co.uk
PING bbc.co.uk (212.58.244.23) 56(84) bytes of data.

--- bbc.co.uk ping statistics ---
20 packets transmitted, 20 received, 0% packet loss, time 19023ms
rtt min/avg/max/mdev = 220.859/239.723/255.551/9.453 ms
[mick@house ~]$
 
R

reactor_sa

Executive Member
Joined
Feb 6, 2009
Messages
7,844
streaker said:
Bandwidth were added to all three breakout points. So from the feedback so far, Durban seems problematic. How are the other areas doing ?
Click to expand...
Performance on the dowerglen msan seems noticeably better,.can actually complete a speedtest.
 
D

Deckert

Well-Known Member
Joined
Jan 13, 2004
Messages
386
Packet-loss to MTN and some intl destination persisting

Hey streaker,

streaker said:
There is clearly an issue on the peering our upstream provider SAIX have with MTN. This was reported to them and we are awaiting resolution.
Click to expand...

Any news back from SAIX yet? Packet-loss is still exactly the same as before, so I'm assuming SAIX has not yet fixed whatever the issue is. From an MTR trace, the trouble begins between hop 6 and 7. Peering link issue perhaps?

Code: 
HOST: spica                       Loss%   Snt   Last   Avg  Best  Wrst StDev
  4.|-- 105.224.0.138              0.0%   100    5.2   5.1   4.4   6.5   0.3
  5.|-- 105.228.0.13               0.0%   100    4.6   4.7   4.5   6.2   0.3
  6.|-- 165.165.216.217            0.0%   100    5.3   5.4   5.2   6.0   0.1
[b]  7.|-- 196.43.25.138             37.0%   100    5.6   5.7   5.3   6.5   0.3 [/b]
  8.|-- 196.25.58.18              37.0%   100    6.7   6.9   6.6   7.2   0.1
  9.|-- 41.181.139.108            37.0%   100    7.1  12.1   7.0  51.9   9.4
 10.|-- 196.44.31.95              33.0%   100    7.5   8.9   7.3  35.6   5.6
 11.|-- 41.181.180.8              35.0%   100    7.4   8.2   7.1  38.4   4.6
 12.|-- 196.30.230.53             41.0%   100    7.4   7.5   7.3   7.9   0.1

And it's really bad. Here's my results to the same location MickZA is pinging to:

--- 212.58.244.23 ping statistics ---
20 packets transmitted, 15 received, 25% packet loss, time 19008ms
rtt min/avg/max/mdev = 256.312/273.207/292.017/11.041 ms

Similarly, tracing to Linode's Dallas DC at The Planet in the US:

Code: 
  4.|-- 105.224.0.6                0.0%   100    4.9   5.0   4.5   6.4   0.2
  5.|-- 105.228.0.13               0.0%   100    4.7   4.8   4.5   6.4   0.3
  6.|-- 165.165.216.217            0.0%   100    5.5   5.4   4.7   6.2   0.2
[b]  7.|-- 196.43.9.50               35.0%   100  168.9 169.2 168.3 173.6   0.7[/b]
  8.|-- 149.6.98.73               42.0%   100  188.2 188.2 187.9 189.5   0.3
  9.|-- 130.117.2.109             27.0%   100  187.7 188.0 187.1 192.6   0.7
 10.|-- 154.54.58.173             35.0%   100  183.5 183.4 182.9 184.1   0.2
 11.|-- 154.54.1.117              34.0%   100  248.3 248.4 248.2 248.7   0.1
 12.|-- 154.54.43.10              34.0%   100  249.6 250.0 249.6 256.8   0.9
 13.|-- 154.54.26.129             24.0%   100  269.5 269.4 268.9 269.9   0.2
 14.|-- 154.54.6.221              35.0%   100  276.7 277.0 276.6 277.4   0.2
 15.|-- 154.54.42.165             34.0%   100  284.0 284.0 283.6 284.5   0.2
 16.|-- 154.54.3.133              36.0%   100  293.8 293.9 293.5 294.8   0.2
 17.|-- 154.54.28.74              46.0%   100  294.2 294.3 293.7 294.7   0.2
 18.|-- ???                       100.0   100    0.0   0.0   0.0   0.0   0.0
 19.|-- 173.255.207.17            33.0%   100  350.5 335.3 305.6 361.5  13.8
 20.|-- 50.116.25.154             30.0%   100  324.8 321.8 293.4 347.8  12.5

Note that, for both cases it is happening on all TI links I supervise, including a 100Mbit fibre link, from which the above tests are taken.

--deckert
 
M

moedawood

Senior Member
Joined
Apr 20, 2016
Messages
910
Code: 
|                       Host              -   %  | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
|      105-187-252-70.ti-sen.telkomsa.net -    2 |   98 |   97 |   16 |   32 |   90 |   37 |
|                            105.229.0.13 -    0 |  102 |  102 |   19 |   33 |   84 |   35 |
|                           105.187.249.6 -    2 |   98 |   97 |   29 |   47 |   84 |   65 |
|                         105.187.249.242 -    0 |  102 |  102 |   30 |   41 |   84 |   35 |
|                   cloudflare.ixp.joburg -    0 |  102 |  102 |   28 |   49 |  664 |   37 |
|                           104.20.10.169 -    2 |   98 |   97 |   28 |   44 |   90 |   67 |

I'm also experiencing a small amount of packet loss.

my trace:

Code: 
traceroute to mybroadband.co.za (104.20.10.169), 30 hops max, 40 byte packets
 1  * * *
 2  105.187.252.70 (105.187.252.70)  14.455 ms  14.974 ms  14.16 ms
 3  105.229.0.13 (105.229.0.13)  14.892 ms  14.852 ms  15.322 ms
 4  105.187.249.6 (105.187.249.6)  26.508 ms  26.013 ms  26.215 ms
 5  105.187.249.6 (105.187.249.6)  25.121 ms  26.638 ms  26.478 ms
 6  cloudflare.ixp.joburg (196.60.8.198)  25.969 ms  25.78 ms  26.444 ms
 7  104.20.10.169 (104.20.10.169)  25.275 ms  26.276 ms  25.998 ms
 
M

moedawood

Senior Member
Joined
Apr 20, 2016
Messages
910
To afrihost's website:

Code: 
|                       Host              -   %  | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|

|105-187-252-70.ti-sen.telkomsa.net -    3 |   47 |   46 |   20 |   51 |  369 |   27 |
|                            105.229.0.13 -    5 |   43 |   41 |   17 |   44 |  242 |   32 |
|                           105.187.249.6 -    5 |   43 |   41 |   29 |   60 |  248 |   69 |
|                           105.187.249.6 -    5 |   43 |   41 |   31 |   60 |  279 |   34 |
|                   cloudflare.ixp.joburg -    3 |   47 |   46 |   30 |   55 |  248 |   51 |
|                           104.20.31.244 -    5 |   43 |   41 |   31 |   57 |  213 |   43 |

trace:

Code: 
traceroute to afrihost.com (104.20.30.244), 30 hops max, 40 byte packets
 1  * * *
 2  105-187-252-70.ti-sen.telkomsa.net (105.187.252.70)  15.868 ms  14.774 ms  14.138 ms
 3  105.229.0.13 (105.229.0.13)  15.153 ms  15.711 ms  14.835 ms
 4  105.187.249.6 (105.187.249.6)  27.205 ms  27.096 ms  25.952 ms
 5  105.187.249.6 (105.187.249.6)  26.405 ms  25.466 ms 105.187.249.242 (105.187.249.242)  25.532 ms
 6  cloudflare.ixp.joburg (196.60.8.198)  24.786 ms  25.996 ms  26.476 ms
 7  104.20.30.244 (104.20.30.244)  26.201 ms  26.444 ms  27.452 ms

Should I be worried?
 
S

streaker

Active Member
Joined
Jan 24, 2008
Messages
47
Deckert said:
Hey streaker,



Any news back from SAIX yet? Packet-loss is still exactly the same as before, so I'm assuming SAIX has not yet fixed whatever the issue is. From an MTR trace, the trouble begins between hop 6 and 7. Peering link issue perhaps?

Code: 
HOST: spica                       Loss%   Snt   Last   Avg  Best  Wrst StDev
  4.|-- 105.224.0.138              0.0%   100    5.2   5.1   4.4   6.5   0.3
  5.|-- 105.228.0.13               0.0%   100    4.6   4.7   4.5   6.2   0.3
  6.|-- 165.165.216.217            0.0%   100    5.3   5.4   5.2   6.0   0.1
[b]  7.|-- 196.43.25.138             37.0%   100    5.6   5.7   5.3   6.5   0.3 [/b]
  8.|-- 196.25.58.18              37.0%   100    6.7   6.9   6.6   7.2   0.1
  9.|-- 41.181.139.108            37.0%   100    7.1  12.1   7.0  51.9   9.4
 10.|-- 196.44.31.95              33.0%   100    7.5   8.9   7.3  35.6   5.6
 11.|-- 41.181.180.8              35.0%   100    7.4   8.2   7.1  38.4   4.6
 12.|-- 196.30.230.53             41.0%   100    7.4   7.5   7.3   7.9   0.1

And it's really bad. Here's my results to the same location MickZA is pinging to:

--- 212.58.244.23 ping statistics ---
20 packets transmitted, 15 received, 25% packet loss, time 19008ms
rtt min/avg/max/mdev = 256.312/273.207/292.017/11.041 ms

Similarly, tracing to Linode's Dallas DC at The Planet in the US:

Code: 
  4.|-- 105.224.0.6                0.0%   100    4.9   5.0   4.5   6.4   0.2
  5.|-- 105.228.0.13               0.0%   100    4.7   4.8   4.5   6.4   0.3
  6.|-- 165.165.216.217            0.0%   100    5.5   5.4   4.7   6.2   0.2
[b]  7.|-- 196.43.9.50               35.0%   100  168.9 169.2 168.3 173.6   0.7[/b]
  8.|-- 149.6.98.73               42.0%   100  188.2 188.2 187.9 189.5   0.3
  9.|-- 130.117.2.109             27.0%   100  187.7 188.0 187.1 192.6   0.7
 10.|-- 154.54.58.173             35.0%   100  183.5 183.4 182.9 184.1   0.2
 11.|-- 154.54.1.117              34.0%   100  248.3 248.4 248.2 248.7   0.1
 12.|-- 154.54.43.10              34.0%   100  249.6 250.0 249.6 256.8   0.9
 13.|-- 154.54.26.129             24.0%   100  269.5 269.4 268.9 269.9   0.2
 14.|-- 154.54.6.221              35.0%   100  276.7 277.0 276.6 277.4   0.2
 15.|-- 154.54.42.165             34.0%   100  284.0 284.0 283.6 284.5   0.2
 16.|-- 154.54.3.133              36.0%   100  293.8 293.9 293.5 294.8   0.2
 17.|-- 154.54.28.74              46.0%   100  294.2 294.3 293.7 294.7   0.2
 18.|-- ???                       100.0   100    0.0   0.0   0.0   0.0   0.0
 19.|-- 173.255.207.17            33.0%   100  350.5 335.3 305.6 361.5  13.8
 20.|-- 50.116.25.154             30.0%   100  324.8 321.8 293.4 347.8  12.5

Note that, for both cases it is happening on all TI links I supervise, including a 100Mbit fibre link, from which the above tests are taken.

--deckert
Click to expand...

deckert,

SAIX has acknowledged the issue but does not provide feedback if the the issue will be resolved. We have already made contact with MTN and they agreed to peer with us at NAPAfrica in Johannesburg and Cape Town. Just waiting on them to complete their configurations.

Will ask them for a progress update again today
 
D

Deckert

Well-Known Member
Joined
Jan 13, 2004
Messages
386
streaker said:
SAIX has acknowledged the issue but does not provide feedback if the the issue will be resolved. We have already made contact with MTN and they agreed to peer with us at NAPAfrica in Johannesburg and Cape Town. Just waiting on them to complete their configurations. Will ask them for a progress update again today
Click to expand...

Thank you, streaker. Really appreciate your effort on this one.

Hopefully SAIX gets things fixed anyway, since the same issue seems to impact the Intl. traffic as well.

--deckert
 
D

Draken

Well-Known Member
Joined
Apr 5, 2009
Messages
386
streaker said:
deckert,

SAIX has acknowledged the issue but does not provide feedback if the the issue will be resolved. We have already made contact with MTN and they agreed to peer with us at NAPAfrica in Johannesburg and Cape Town. Just waiting on them to complete their configurations.

Will ask them for a progress update again today
Click to expand...


Not sure what you mean about 'if the issue will be resolved'.

There is no IF, people pay for a service, and they haven't been getting it pretty much this whole month. Any normal company in any normal country would be have it fixed it hours, and have refunded the customer for wasted time.

SA, where you pay 10x the price of every other country for 10x ****ter internet. Imagine our ISPS were run by people educated on this matter :(
 
ponder

ponder

Honorary Master
Joined
Jan 22, 2005
Messages
92,823
Draken said:
Not sure what you mean about 'if the issue will be resolved'.

There is no IF, people pay for a service, and they haven't been getting it pretty much this whole month. Any normal company in any normal country would be have it fixed it hours, and have refunded the customer for wasted time.

SA, where you pay 10x the price of every other country for 10x ****ter internet. Imagine our ISPS were run by people educated on this matter :(
Click to expand...

He's saying that when there is a resolution in the pipeline they don't provide feedback, not if they're going to fix it or not. SA, where kids don't pay attention in English class during comprehension lessons :p
 
D

Draken

Well-Known Member
Joined
Apr 5, 2009
Messages
386
ponder said:
He's saying that when there is a resolution in the pipeline they don't provide feedback, not if they're going to fix it or not. SA, where kids don't pay attention in English class during comprehension lessons :p
Click to expand...

Keep supporting and defending **** service, our country will go so far by doing so.

Main issue is from PANORAMA Exchange(196.25.166.225)

Ping statistics for 196.25.166.225:
Packets: Sent = 186, Received = 150, Lost = 36 (19% loss)

It happens every couple of years when a lot of people suddenly jump up tiers of connection speeds, from good deals or just natural line speed increases.

There is not even any faults reported for it on the notice board, seems like telkom really has under 'control', good lols.
 
Last edited:
You must log in or register to reply here.
Top