Saol 30 Gig slow?

do traceroute to www.is.co.za (cant ping this, icmp blocked) or news.is.co.za

ps. if its data pooling their doing, then its not a teething problem but simply the way they run the network(easiest way to cut out heavy constant downloading users) and something they have in place for just that.

pss. I see mtr's to saol.com shows very irratic latency between 50-200ms spikes, looks alot like congestion. Test done from 3 different server locations ie. IS/UUNET/SAIX to saol.com(presumably saol.com is hosted on their own ipc?)
 
Last edited:
Wont let me trace em :confused:

However ping to news is

Pinging news.is.co.za [168.209.98.66] with 32 bytes of data:

Reply from 168.209.98.66: bytes=32 time=531ms TTL=237
Reply from 168.209.98.66: bytes=32 time=541ms TTL=237
Reply from 168.209.98.66: bytes=32 time=510ms TTL=237
Reply from 168.209.98.66: bytes=32 time=554ms TTL=237

Ping statistics for 168.209.98.66:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 510ms, Maximum = 554ms, Average = 534ms
 
Traceroutes should work fine, ping is normal tho.. any network outside IS would ping high to the news.is.co.za server

check some ping test to:

kalahari.mweb.co.za
karoo.mweb.co.za
mweb.co.za
 
Yikes


Pinging karoo.mweb.co.za [196.2.128.100] with 32 bytes of data:

Reply from 196.2.128.100: bytes=32 time=276ms TTL=111
Reply from 196.2.128.100: bytes=32 time=320ms TTL=111
Reply from 196.2.128.100: bytes=32 time=395ms TTL=111
Reply from 196.2.128.100: bytes=32 time=403ms TTL=111

Ping statistics for 196.2.128.100:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 276ms, Maximum = 403ms, Average = 348ms



Pinging mweb.co.za [196.2.63.110] with 32 bytes of data:

Request timed out.
Reply from 196.2.63.110: bytes=32 time=596ms TTL=32
Reply from 196.2.63.110: bytes=32 time=477ms TTL=32
Request timed out.

Ping statistics for 196.2.63.110:
Packets: Sent = 4, Received = 2, Lost = 2 (50% loss),
Approximate round trip times in milli-seconds:
Minimum = 477ms, Maximum = 596ms, Average = 536ms


Kalahari failed :(
 
Now the speed pushed up again...

Pinging mweb.co.za [196.2.63.110] with 32 bytes of data:

Reply from 196.2.63.110: bytes=32 time=98ms TTL=32
Reply from 196.2.63.110: bytes=32 time=114ms TTL=32
Reply from 196.2.63.110: bytes=32 time=97ms TTL=32
Reply from 196.2.63.110: bytes=32 time=106ms TTL=32

Ping statistics for 196.2.63.110:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 97ms, Maximum = 114ms, Average = 103ms

Will drop soon i imagine to bad :(

Your line speed is approximately 441.1 Kbps or 54.1 kBytes/sec
( Where kb = kilobits and kB = kiloBytes

Download time: 9.25 seconds
Size of file: 500 KiloBytes
Estimated line speed: 441.1 (kilobits/second)
Estimated line speed: 54.1 (kiloBytes/second)


Pinging saix.net [196.25.1.200] with 32 bytes of data:

Reply from 196.25.1.200: bytes=32 time=69ms TTL=243
Reply from 196.25.1.200: bytes=32 time=81ms TTL=243
Reply from 196.25.1.200: bytes=32 time=74ms TTL=243
Reply from 196.25.1.200: bytes=32 time=74ms TTL=243

Ping statistics for 196.25.1.200:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 69ms, Maximum = 81ms, Average = 74ms
 
Last edited:
I dont know for sure yet, but it definitely seems like their doing some sort of pooling. Thats why the bursts in speed happen.

I would suggest to saol to only implement something like this on specific protocols/port combinations, leaving it disable for latency intensive services like gaming :/

Saol care to comment on this? Would be interesting to know your view on the matter.
 
Yeah i would love to know too :( and how some people get high speeds constantly :( over IPC

Tried P2P over PayVPN seems to pull barely over 20kb/s too.(10 gig shaped SAIX was 55kb/s) How does IPC shape?

Sent an email to Marko and to Support again with all the info attached here Pings etc etc
 
Last edited:
Borgir offered me his account to test out some things, below the dreaded results o_0

saix.net:

Pinging saix.net [196.25.1.200] with 32 bytes of data:

Reply from 196.25.1.200: bytes=32 time=398ms TTL=244
Reply from 196.25.1.200: bytes=32 time=390ms TTL=244
Reply from 196.25.1.200: bytes=32 time=378ms TTL=244
Reply from 196.25.1.200: bytes=32 time=386ms TTL=244
Reply from 196.25.1.200: bytes=32 time=416ms TTL=244
Reply from 196.25.1.200: bytes=32 time=414ms TTL=244
Reply from 196.25.1.200: bytes=32 time=442ms TTL=244
Reply from 196.25.1.200: bytes=32 time=449ms TTL=244
Reply from 196.25.1.200: bytes=32 time=520ms TTL=244

Ping statistics for 196.25.1.200:
Packets: Sent = 9, Received = 9, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 378ms, Maximum = 520ms, Average = 421ms


mweb.co.za:

Pinging mweb.co.za [196.2.63.110] with 32 bytes of data:

Reply from 196.2.63.110: bytes=32 time=453ms TTL=98
Reply from 196.2.63.110: bytes=32 time=415ms TTL=98
Reply from 196.2.63.110: bytes=32 time=400ms TTL=98
Request timed out.
Request timed out.
Reply from 196.2.63.110: bytes=32 time=571ms TTL=98
Request timed out.
Reply from 196.2.63.110: bytes=32 time=434ms TTL=98
Reply from 196.2.63.110: bytes=32 time=441ms TTL=98

Ping statistics for 196.2.63.110:
Packets: Sent = 9, Received = 6, Lost = 3 (33% loss),
Approximate round trip times in milli-seconds:
Minimum = 400ms, Maximum = 571ms, Average = 301ms


cheetah-is.co.za:

At this point all local access died but I could still do international tests.



google.com:

Pinging google.com [64.233.187.99] with 32 bytes of data:

Reply from 64.233.187.99: bytes=32 time=775ms TTL=239
Reply from 64.233.187.99: bytes=32 time=680ms TTL=239
Reply from 64.233.187.99: bytes=32 time=718ms TTL=239
Reply from 64.233.187.99: bytes=32 time=718ms TTL=239

Ping statistics for 64.233.187.99:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 680ms, Maximum = 775ms, Average = 722ms


cs.jolt.co.uk:

Pinging cs.jolt.co.uk [195.149.21.9] with 32 bytes of data:

Reply from 195.149.21.9: bytes=32 time=611ms TTL=52
Reply from 195.149.21.9: bytes=32 time=588ms TTL=52
Reply from 195.149.21.9: bytes=32 time=603ms TTL=52
Reply from 195.149.21.9: bytes=32 time=582ms TTL=52

Ping statistics for 195.149.21.9:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 582ms, Maximum = 611ms, Average = 596ms


Last but not least, saol.com:

Pinging saol.com [196.25.188.9] with 32 bytes of data:

Reply from 196.25.188.9: bytes=32 time=352ms TTL=124
Reply from 196.25.188.9: bytes=32 time=329ms TTL=124
Reply from 196.25.188.9: bytes=32 time=328ms TTL=124
Reply from 196.25.188.9: bytes=32 time=382ms TTL=124

Ping statistics for 196.25.188.9:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 328ms, Maximum = 382ms, Average = 347ms



As you can see from the tests, the latency is absolutely screwed, either by shaping methods used or congestion, SAOL care to comment ? Id put my money on congestion causing these high(local even) latency.

Tracing route to saol.com [196.25.188.9]
over a maximum of 30 hops:

1 9 ms 10 ms 11 ms 196.41.110.1
2 371 ms 41 ms 371 ms saol-ipc-wan1-saol.saol.com [196.41.96.2]
3 377 ms 405 ms 430 ms www.saol.com [196.25.188.9]

The high latency spike definitely happens as soon as you enter the SAOL IPC, interestingly enough, is SAOL offering the IPC over a wireless access point(saving money this way hooked wireless to some datacentre?) or is that -wan1- entry in the dns just coincidence.


Tracing route to google.com [64.233.187.99]
over a maximum of 30 hops:

1 10 ms 10 ms 10 ms 196.41.110.1
2 353 ms 283 ms 343 ms saol-ipc-wan1-saol.saol.com [196.41.96.2]
3 216 ms 191 ms 246 ms astroflowguard.saol.com [196.25.188.3]
4 266 ms 350 ms 287 ms core-access-rtr.saol.com [196.25.188.1]
5 314 ms 328 ms 306 ms 196.41.96.254
6 349 ms 321 ms 328 ms 196.15.203.73
7 * * * Request timed out.
8 572 ms 570 ms 559 ms ny-ip-dir-globalc-pos-7-0.telkom-ipnet.co.za [19
6.43.9.149]
9 599 ms * 518 ms g0-2.na01.b001105-24.jfk02.atlas.cogentco.com [3
8.112.18.41]
10 426 ms 519 ms 830 ms g10-2-103.core01.jfk02.atlas.cogentco.com [38.11
2.38.165]
11 609 ms * 517 ms p12-0.core01.jfk05.atlas.cogentco.com [154.54.1.
26]
12 581 ms 545 ms 594 ms core1-0-0-8.lga.net.google.com [198.32.118.39]
13 661 ms 548 ms 523 ms 66.249.94.235
14 600 ms 639 ms 505 ms 72.14.238.232
15 545 ms 562 ms 654 ms 66.249.95.149
16 * 564 ms 631 ms 72.14.236.173
17 654 ms 668 ms 612 ms 216.239.49.222
18 615 ms 623 ms 594 ms 64.233.187.99

Trace complete.


Super-high and lossy connection to google.com.


So now I tried a saix account ping test to the saol ipc network and I get the following:

196.41.110.1:

Pinging 196.41.110.1 with 32 bytes of data:

Reply from 196.41.110.1: bytes=32 time=359ms TTL=255
Reply from 196.41.110.1: bytes=32 time=319ms TTL=255
Reply from 196.41.110.1: bytes=32 time=314ms TTL=255
Reply from 196.41.110.1: bytes=32 time=281ms TTL=255

Ping statistics for 196.41.110.1:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 281ms, Maximum = 359ms, Average = 318ms

saol-ipc-wan1-saol.saol.com:

Pinging saol-ipc-wan1-saol.saol.com [196.41.96.2] with 32 bytes of data:

Reply from 196.41.96.2: bytes=32 time=67ms TTL=248
Reply from 196.41.96.2: bytes=32 time=75ms TTL=248
Reply from 196.41.96.2: bytes=32 time=58ms TTL=248
Reply from 196.41.96.2: bytes=32 time=44ms TTL=248

Ping statistics for 196.41.96.2:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 44ms, Maximum = 75ms, Average = 61ms


Looking at the results, the congestion is definitely between the 196.41.110.1 and saol-ipc-wan1-saol.saol.com peering point, congestion? or wireless link?

All speculation aside, I dont want to critise anyone here, just would like some answers to my questions from a knowledgable SAOL employee, nothing more, nothing less.

ps. If above doesnt make sense, feel free to ask me to explain it again.
 
If your speculation is true and they are using wireless connection, how much influence does the weather have cause it has been a real sh1tty day here in durbs, rain and heavily overcast?
 
Thats the point Im trying to make, Wireless is unstable, add some weather and it would be even worse.

Now if SAOL could clarify and state, we do use wireless and during bad weather it will cause our main peering link to perform slow, then everyone know where they stand.

ps. This is still speculation, dont take my word for it.
 
Awesome post and Thanks Clipse.


Point is it is pathetic. First day was fine and gradually getting worse. I noticed no improvement since others claim teething issues have been fixed.

It is looking more a waste of R599 since at this rate it is unusable. I hope this also clears that since Clipse tested and tried account it is NOT a Telkom issue my side before being told that it is :mad: I cant see how people on this get 122kb/s +- Dl speeds.
 
Last edited:
I had a very bad experince with SAOL IPC . Clipse and Borgir confirm my feelings. On the other hand other users reported very good speeds and other ISPs are using it. The question is , Is it SAOL problem or the technology problem ?
 
Slow IPC access

Hi everybody,

Now I would like to give you all the actual facts about SAOL’s IPC.
I am posting here to let our clients know what is going on and not to get into an argument or vindictive posting frenzy.

The facts are as such:

We are not supplying IPC bandwidth via wireless; we are using a fiber optic link to our Martis node onsite supplies by Telkom from the exchange.
Now we load balance over multiple 2Meg links to the IPC network and the problem here is that Telkom is having problems doing per packet load balancing over those links (this is something we are working on to sort the problem out) and as such we are doing per connection load balancing, now what happens here is if you send a ping and it goes via 2Meg link number 1 and that link is congested, well you know the result.
We are in the process of sorting this out with Telkom but as you all know we are in the Telkom FREEZE period now and we are battling to get them to do anything.
For the record, yes we do bandwidth management on the IPC network to allow an equal distribution of bandwidth to all, not allowing one client to consume all the bandwidth. We also do NOT allow any p2p on our links as this would simply consume all our EXPENSIVE bandwidth and make it impossible to service our clients.

To Sum this up, we are not having teething problems, we are having Telkom problems and are working on this problem full time to sort it out.

I hope this answers your questions, I will respond to any legit posts regarding this matter. If you want to have a go at SAOL, please don't be surprised if you get no answer.

Tarquin
 
Last edited:
Hi Tarquin

Thanks for the reply, all I needed to hear.

Hope you guys get things sorted.
 
Tarquin said:
If you want to have a go at SAOL, please don't be surprised if you get no answer.

Tarquin
Click to expand...

Wrong attitude mate, sort it.

/doc
 
Thanks.

Lets hope it gets resolved quickly =) As it is really unstable at mo :(
 
Last edited:
This is the type of an answer which I am expecting from ISP.
 
Apart from saying...Look...Sorry everyone who paid the R599 will get a 10 gig account + whatever else to compensate till the issue is fixed ...They dont.

So we cant say this...cause they will class it an attack on them.

Instead they say sorry your stuck on this terrible line that barely works for no ETA but keep sending us the R599. Why of WHY release a product that is messed up and not compensate
 
Well it is good that they explain the issue, but this bit is pretty aggressive:

Tarquin said:
If you want to have a go at SAOL, please don't be surprised if you get no answer.
Click to expand...

:(
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X