Congestion: Eldoraigne, Centurion (012 653 xxxx)

Deckert

Deckert

Well-Known Member
Joined
Jan 13, 2004
Messages
425
Reaction score
38
Location
Centurion, South Africa
Hi,

[Edit: This issue has been resolved, see update at the end of this thread]

Agreed with other posters here - congestion is real and happening. This post is for the benefit of Telkom.

This shows the latency of the Eldoraigne (012 653) exchange for the past 10 days. The line is unloaded. The latency starts at around 5-6pm and usually abates by around 11:30pm. I measure peak rates of around 300-500ms over my line. Note that during this time, my sync rates are stable and error counts are zero. I have indicated two large youtube uploads with green arrows (so that's acceptable). The red arrows show plain old exchange congestion.

latency-10days.png

Here is the typical recovery curve for this exchange. Note that the IP I am pinging is the Telkom 1st-hop on the other side of my DSL link (105.224.216.1 in my case) so no other factors are causing the latency.

latency-3hrs.png

Please, Telkom - you have to know this Exchange is under severe load.

--deckert
 
Last edited:
Severe latency issues here: 012 653 XXXX Eldoraigne X18

I phones Telkom support last night, they degraded my 10mbps line to 7mbps, NO EFFECT.

But this morning I was able to browse perfectly with sub 20ms.

What you need to do is create a tracert to "saix.net" and take a few speedtest.net screenshots and post them to TelkomZA.

They seem to believe nothing is faulty.

I get between 400-600ms to saix.net and the local midrand speedtest server.
 
Hi Neighbours

I am on the same exchange, 012 653 xxxx

This is the feedback I received from TelkomZA.
But I still need to send him some traceroutes.

Had a look and unable to see any congestion on the DSLam, peaks at just over 150mbps and link is set to 240mbps
Click to expand...
 
vaalpens said:
Hi Neighbours

I am on the same exchange, 012 653 xxxx

This is the feedback I received from TelkomZA.
But I still need to send him some traceroutes.
Click to expand...

Yes this is what they tell me as well, the problem is that they test the exchange when there isnt any congestion. If they check the dslam from any between 5pm and 4am they will see it.

My SNR is usually in the 13db, when my line goes haywire it gets reduced to as much as 5db.

Wierd but i think they are in denial about the exchange giving problems.

EDIT: I have another buddy of mine on the same exchange, same issues at exactly the same time. We are compiling a tracert and latency test for TelkomZA

Capture.jpg

[-]Apologies for the poor formatting[/-]

My speedtest.net results from my house to the midrand server.

There results are quite indicative that something serious is wrong (check the latency)

This was a 10mbps product that was working fine till mid January

Here is my (Still ongoing) previous thread: http://mybroadband.co.za/vb/showthr...-Wierda-Park-Area-Latency-Problems?highlight=
 
Last edited:
Eldoraigne exchange as well but a 012 654 xxxx number and exactly the same issues.
 
Hi guys,

This evening was a little better compared to yesterday-evening. I suspect that Wednesdays might be lighter loaded in general. Still, some increase in latency. As always, the line these tests are done from is completely unloaded:

latency-1day.png

Please indicate (in this thread) if you have a similar problem.

--deckert
 
This weekend has been particularly horrible, 500-600ms.

Since the technicians keep closing my ticket without even a phone call, I have decided to repeatedly open a new ticket and harass them about this issue:

photo.PNG
 
They did a port reset on my line now it is unable to sync to any ISP's, Telkom keeps closing the tickets and refuse to see a issue because the line is syncing. I have tested my router on another line and works fine. I know Nether had the same issue the exchange seems extremely unstable
 
Splendid said:
They did a port reset on my line now it is unable to sync to any ISP's, Telkom keeps closing the tickets and refuse to see a issue because the line is syncing. I have tested my router on another line and works fine. I know Nether had the same issue the exchange seems extremely unstable
Click to expand...

Finally have internet again but latency's are high again, The DSLAM seems very unstable, when doing a port reset kills your line something is wrong:

see speed test below to rosebank server:

speedtest.PNG
 
Also experiencing congestion in Faerie Glen, with packet loss after hours and over weekends. It has been 2 weeks now :/
 
Hi guys,

Telkom is aware of the situation. Feedback that I got (today) is that, due to many users reporting the same issue, it has been escalated to the area manager for the Hennopspark exchange and that we will hopefully get some feedback from him/her. I am confident that Telkom will be able to sort this out.

--deckert
 
Hi,

Just for reference, here are the latency/times for the evening of the 17th (i.e. last night). Latency peaked at 700ms at 10pm, but I only took the traces form 11pm onwards. As you can see, the latency recovered completely at 00:35. As before, the link was completely unloaded.

traceroute-1day-20140217.png

--deckert
 
Update for 17-18 of Feb

The evening latency was just absolutely crazy. It started off mild at 200ms, but escalated to around 800ms!

photo.PNG

Once it was morning again on the 18 feb latency was perfect and I was able to achieve high speeds.
 
Sorry to hear the problems, I too was using this exchange in the past and it would congest often and they would add capacity every time after a moerse fight - Fortunately i'm no longer using this exchange. If you search these forums you will find similar threads for this exchange.
 
bleh, down from 10Mbps sync to 1422 KB/s :'(
latency is fine now, but wait till 7pm... :(

I also have the issue where if you do a port reset via AfriHost's clientzone, I can't establish PPPOE connections anymore, and have to call Telkom to do a port reset.
 
Last edited:
A Telkom representative called me yesterday to esquire about the issue, the major problem is that she doesn't understand what latency is, I asked her politely to refer the problem to the technician, he has to call me.

If they close my ticket without so much as a phone call from them I will just reopen it repeatedly. And unfortunately if they still refuse to address the issue I will have to cancel my Telkom line, its basically useless anyway.

Downgrading my line has had no effect.

I have also seriously considered reporting this to the NCC, false advertisements and promises that don't reflect the poor quality of their network.
 
Hi again,

Here's an update of what happened last night in terms of latency on the Eldoraigne (Hennopspark) exchange. Note the "outage" where we had the suburb-wide power failure ... when the power came back on, almost immediately after that the latency was fine, but as people started connecting, latency increased again. Last night's latency was some of the worst I've seen:

latency-1day-20140218.png

and a traceroute:

Code: 
root@zorg:~# mtr -c10 -r -n google.com
HOST: zorg                        Loss%   Snt   Last   Avg  Best  Wrst StDev
  1.|-- 192.168.1.1                0.0%    10    0.5   0.5   0.4   0.7   0.1
  2.|-- [b]105.224.216.1[/b]              0.0%    10  [b]795.1 726.6 682.2 795.1[/b]  40.3
  3.|-- 105.224.0.6               10.0%    10  785.6 727.7 691.6 789.7  38.0
  4.|-- 105.224.0.13              10.0%    10  784.9 727.8 675.3 810.6  48.3
  5.|-- 196.25.39.37              10.0%    10  783.6 729.4 677.9 799.1  44.6
  6.|-- 196.43.11.134             10.0%    10  780.6 725.0 663.4 780.6  44.6
  7.|-- 196.43.26.106             10.0%    10  784.5 724.4 660.9 784.5  43.4
  8.|-- 72.14.212.1               10.0%    10  786.9 728.6 662.2 786.9  42.0
  9.|-- 64.233.174.21             11.1%     9  807.1 735.4 657.0 807.1  46.6
 10.|-- 74.125.233.34             11.1%     9  808.0 735.2 646.9 808.0  51.5

--deckert
 
Deckert said:
Hi again,

Here's an update of what happened last night in terms of latency on the Eldoraigne (Hennopspark) exchange. Note the "outage" where we had the suburb-wide power failure ... when the power came back on, almost immediately after that the latency was fine, but as people started connecting, latency increased again. Last night's latency was some of the worst I've seen:

View attachment 100753

and a traceroute:

Code: 
root@zorg:~# mtr -c10 -r -n google.com
HOST: zorg                        Loss%   Snt   Last   Avg  Best  Wrst StDev
  1.|-- 192.168.1.1                0.0%    10    0.5   0.5   0.4   0.7   0.1
  2.|-- [b]105.224.216.1[/b]              0.0%    10  [b]795.1 726.6 682.2 795.1[/b]  40.3
  3.|-- 105.224.0.6               10.0%    10  785.6 727.7 691.6 789.7  38.0
  4.|-- 105.224.0.13              10.0%    10  784.9 727.8 675.3 810.6  48.3
  5.|-- 196.25.39.37              10.0%    10  783.6 729.4 677.9 799.1  44.6
  6.|-- 196.43.11.134             10.0%    10  780.6 725.0 663.4 780.6  44.6
  7.|-- 196.43.26.106             10.0%    10  784.5 724.4 660.9 784.5  43.4
  8.|-- 72.14.212.1               10.0%    10  786.9 728.6 662.2 786.9  42.0
  9.|-- 64.233.174.21             11.1%     9  807.1 735.4 657.0 807.1  46.6
 10.|-- 74.125.233.34             11.1%     9  808.0 735.2 646.9 808.0  51.5

--deckert
Click to expand...

Is there even any point in posting about the issues? Send a message to TelkomZA but no reply yet, send them the link on there twitter channel no response yet and the call desk can only tell me a technician is assigned but nothing is happening this is truly and endless battle just to get what you pay for. Like Nether said it seems that they don't understand what the issue is and what latency means.
 
yep, it is important, the more people we have that complain about the specific exchange, the more information we have to convince them it is not an individual user fault, so everyone on this exchange , please post and complain here
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X