Afrihost - Pure Fibre Feedback Thread

Status
Not open for further replies.
blunt said:
somethings broken with CF it seems

Code: 
traceroute to 1.1.1.1 (1.1.1.1), 20 hops max, 52 byte packets
1  gateway.ibox.za.net (10.23.23.1)  0.665 ms  0.440 ms  0.429 ms
2  * * *
3  165-73-14-113.ip.afrihost.co.za (165.73.14.113)  3.647 ms  3.265 ms  4.458 ms
4  169-1-21-148.ip.afrihost.co.za (169.1.21.148)  3.788 ms  5.549 ms  4.237 ms
5  169-1-21-132.ip.afrihost.co.za (169.1.21.132)  3.446 ms
    169-1-21-133.ip.afrihost.co.za (169.1.21.133)  3.860 ms  5.958 ms
6  * * 169-1-21-133.ip.afrihost.co.za (169.1.21.133)  9.050 ms
7  * * *
8  one.one.one.one (1.1.1.1)  320.875 ms  318.953 ms  281.920 ms
Click to expand...

Cloudflare Outages in ZA and parts of Africa.

CPT
DUR
Kigali
Some of JHB traffic rerouted

So CPT and DUR goes to JHB but some traffic from JHB rerouted to EU or Dijbouti possibly due to congestion.
 
waylander said:
Outages on Cloudflare means outages on Seacom(?)

Saw CISP thread complains too, possibly something happening up in north Africa
Click to expand...

Doesn't look like it. Both routes via Seacom and WACS looks good.

seacom_seacom.png

seacom_wacs.png
 
Hi Folks

Can someone please give me an idea of what is going on here. Every evening, like clock work, my connection degrades with speeds being lower and erratic this is usually coupled with higher latency and ping as well as packet loss. I have had Vuma test the line and they could not find anything wrong with the line. I have also been reliably told from someone inside Vuma that there significant capacity so it cannot be that there network is congested, as suggested before. I ran these same test now this morning and everything is perfectly fine. Last night I had a friend also with Afrihost do the same tests and he had no issues.
 

Attachments

  • Ping to Xbox 2.JPG
    Ping to Xbox 2.JPG
    34.6 KB · Views: 26
  • ping to Xbox 3.JPG
    ping to Xbox 3.JPG
    70.4 KB · Views: 25
  • Ping to Xbox.JPG
    Ping to Xbox.JPG
    86 KB · Views: 24
  • Friends Microsoft Tracert.jpeg
    Friends Microsoft Tracert.jpeg
    102.6 KB · Views: 21
  • My Microsoft Tracert.jpeg
    My Microsoft Tracert.jpeg
    93.4 KB · Views: 21
jdido87 said:
Hi Folks

Can someone please give me an idea of what is going on here. Every evening, like clock work, my connection degrades with speeds being lower and erratic this is usually coupled with higher latency and ping as well as packet loss. I have had Vuma test the line and they could not find anything wrong with the line. I have also been reliably told from someone inside Vuma that there significant capacity so it cannot be that there network is congested, as suggested before. I ran these same test now this morning and everything is perfectly fine. Last night I had a friend also with Afrihost do the same tests and he had no issues.
Click to expand...

This is an issue with the Mweb hosted cache. As you can see from your trace route at hop 5 the traffic is peered with Optinet at Teraco CPT and then the rest of the route is on Optinet/Mwebs network. If there is congestion issues there it's on their side.

My Microsoft Tracert.jpeg
 
waylander said:
Vox
Code: 
 1  * * *
2  41.193.164.85  4.185 ms  3.733 ms  3.517 ms
3  41.193.164.86  3.485 ms  3.243 ms  3.380 ms
4  41.193.164.92  4.516 ms  3.631 ms  3.730 ms
5  196.60.8.120  4.190 ms  4.049 ms  3.998 ms
6  197.80.4.96  5.249 ms  5.128 ms  5.043 ms
7  197.80.5.61  5.432 ms  5.181 ms  5.303 ms
8  197.80.4.225  4.945 ms  4.993 ms  4.898 ms
9  197.80.4.36  4.368 ms  4.803 ms  4.477 ms
10  197.80.7.13  29.573 ms  4.972 ms  5.388 ms
11  2.17.205.224  4.362 ms  4.182 ms  4.208 ms

Afrihost
Code: 
 1  * * *
2  165.73.14.17  5.190 ms  4.055 ms  4.111 ms
3  169.1.21.232  3.734 ms  3.212 ms  3.270 ms
4  169.1.21.233  3.457 ms  3.268 ms  3.200 ms
5  196.60.8.145  3.824 ms  3.593 ms  3.588 ms
6  197.80.4.23  198.713 ms  183.996 ms  182.095 ms
7  197.80.5.57  169.992 ms  170.024 ms  170.285 ms
8  197.80.7.77  192.828 ms  199.186 ms  199.440 ms
9  2.17.205.224  198.278 ms  188.983 ms  181.773 ms
Click to expand...

The issue needs to escalated to Optinet NOC.

5. 196.60.8.145

Is the exchange with Optinet/Mweb. After that they take the traffic to EU and back to ZA to the cache server. I'm wondering why because Afrihost has their own Akamai cache why they are making use of another provider. Vox uses their own cache as you can see.
 
waylander said:
They have their own cache? I thought they just peered with them. TIL
Click to expand...

Either way.

Btw your trace above is traced to the ip address. Do a trace to www.microsoft.com and see it will be local. The Akamai hostname is the same served from a different IP. There is probably just a routing issue to Akamai for CPT. It shouldn't be pulling it from Optinet.
 
cavedog said:
This is an issue with the Mweb hosted cache. As you can see from your trace route at hop 5 the traffic is peered with Optinet at Teraco CPT and then the rest of the route is on Optinet/Mwebs network. If there is congestion issues there it's on their side.

View attachment 836737
Click to expand...
Thank you very much for you help. @cavedog

@AfriGuy can this be addressed?
 
Afrihost/FF
Tracing route to e13678.dspb.akamaiedge.net [2.21.100.214]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms 192.168.1.1
2 1 ms <1 ms <1 ms dlinkrouter.Dlink [192.168.0.1]
3 8 ms 2 ms 3 ms 102-182-162-1.ip.airmobile.co.za [102.182.162.1]
4 3 ms 2 ms 3 ms 169-1-21-91.ip.afrihost.co.za [169.1.21.91]
5 10 ms 12 ms 5 ms akamai.ixp.joburg [196.60.8.212]
6 3 ms 3 ms 3 ms a2-21-100-214.deploy.static.akamaitechnologies.com [2.21.100.214]

Trace complete.

Afrihost/OS
Tracing route to www.microsoft.com [2.21.100.214]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms dlinkrouter [192.168.0.1]
2 * * * Request timed out.
3 7 ms 4 ms 4 ms 165 [165.73.14.17]
4 4 ms 3 ms 4 ms 169 [169.1.21.232]
5 4 ms 3 ms 4 ms 169 [169.1.21.233]
6 10 ms 7 ms 7 ms akamai.ixp.joburg [196.60.8.212]
7 3 ms 3 ms 3 ms www.microsoft.com [2.21.100.214]
 
Moshidi said:
You're not alone :(

Edit: MTN also performing dismissally now, hope it's not another cable break

Edit 2: Back up and running
Click to expand...
Same here on AH/FF. AH/OS is fine

London
1589006361701.png
Midrand
1589006471032.png

/Update

Seems we're back in business. Not used to having to reboot ONT to resolve issues :)
Midrand
1589008268911.png

Hopefully this is not a common occurence..
 
Last edited:
  • Love
Reactions: Yuu
My AH on FF was also broken since just around 4am according to smokeping, speedtests were similar to yours and seems to have been fixed at around 9:25am after an ONT reboot (smokepings haven't run next loop yet in the below pictures). Numerous router reboots didn't help and the ONT reboot seemed to work too.

1589009526870.png
1589009589077.png
 
Last edited:
blunt said:
You OS been upgraded? I got the email from AH about the upgrades but I'm still on 40/20 and 10 May almost over.. guess it's not gonna happen because I downgraded
Click to expand...
Nope, still on 10/5 it seems. Technically we were already on the downgraded packages 1st of May , so the upgrades should've been applied. It's not like everyone was upgraded at 00:01 of the 1st
 
Anyway to change who use the Vumatel Portal for a CPE?
Do I just cancel on my end for Afrihost?
Moving and need to hand over to new "tenant"
 
Pho3nix said:
Anyway to change who use the Vumatel Portal for a CPE?
Do I just cancel on my end for Afrihost?
Moving and need to hand over to new "tenant"
Click to expand...
Yes, all you need to do is cancel the package or email support to move it to the tenant's name.
Please may you kindly cc the new tenant if possible?
 
jdido87 said:
Hi Folks

Can someone please give me an idea of what is going on here. Every evening, like clock work, my connection degrades with speeds being lower and erratic this is usually coupled with higher latency and ping as well as packet loss. I have had Vuma test the line and they could not find anything wrong with the line. I have also been reliably told from someone inside Vuma that there significant capacity so it cannot be that there network is congested, as suggested before. I ran these same test now this morning and everything is perfectly fine. Last night I had a friend also with Afrihost do the same tests and he had no issues.
Click to expand...
Please inbox us so we can get to the bottom of this. Include this thread.
 
Status
Not open for further replies.
Back
Top
Sign up to the MyBroadband newsletter
X