Big Afrihost ADSL problems

Just been talking about how bad things are currently with one of my customers that I persuaded to go with Afrihost. I have stuck with Afrihost a long time now, but getting tired of the issues that crop up. I have always recommended Capped accounts as most don't use more than 50Gb a month. One of my customers switched over to Telkom on Friday and immediately saw an increase in internet speeds. Really considering changing to another ISP. Tired of the speed issues.

Bottom line: capped accounts should not be limited / restricted.

Thanks for the previous recommendations, we really do appreciate it!
Capped accounts are most definitely unshaped and will run as fast as the line can - unfortunately we've seen some technical issues that are causing some slow downs though.

Once resolved things will return to normal.
 
I've been having endless problems with Afrihost - recently - but the last few days have been really bad. Kids are on holiday now and complaining all the time. It keeps giving errors when connecting to sites. I even had to reload this page just now to connect to it.

Really sorry about the issues :( We're seeing a problem on a Google node that our team are working on. Once fixed up you should see things return to normal.
 
AfriGuy,

How about you guys come clean about what really happened a few months ago... leading up to this problem?

Post from NewsHost

Ok well, I've got the 40Mb/s VDSL Businesss Uncapped with them... Was running smoothly doing around 7-8TB a month since December last year... Since beginning April I started getting **** speeds... Phoned them, and had this BS story spun to me from the techie that it’s a Telkom problem, had Telkom investigate for weeks on end at the exchange everything (thanks to a friend of mine)... Could find nothing wrong... 100m away from the MSAN.

So I started using my two test accounts with MWEB and Web Africa... Bam running at max speeds no time outs nothing. Took screen shots of everything mailed it to them. Had a supervisor phoning telling me it’s a problem with MTN... Yet another bull**** story... He told me they would put me into another IP range block... Which helped jack ****...

On Monday I decided to pop into their offices in Rivonia drive and requested to speak to a technical manager regarding my issue. Well, well, well the answer to my complaint was a bit shocking...

We all remember the major MTN Business Gallo Manor datacentre outage in the beginning of the month, well that was just the last straw on the camel's back for them. They use a company in Germany to manage hardware/software locally in Gallo Manor and internationally for them which deals with the prioritizing of all there bandwidth, what ports to block etc etc, you get the picture. Now the system started crashing before the Gallo Manor incident, but when that happened everything went to hell... Ever since then, they have tried everything to get the system back up and running... NO proper backups in place to roll back to nothing... I was informed that they are looking to having everything back to normal only in late June early July....

This was confirmed by a technical manager from Afrihost
 
We all remember the major MTN Business Gallo Manor datacentre outage in the beginning of the month, well that was just the last straw on the camel's back for them. They use a company in Germany to manage hardware/software locally in Gallo Manor and internationally for them which deals with the prioritizing of all there bandwidth, what ports to block etc etc, you get the picture. Now the system started crashing before the Gallo Manor incident, but when that happened everything went to hell... Ever since then, they have tried everything to get the system back up and running... NO proper backups in place to roll back to nothing... I was informed that they are looking to having everything back to normal only in late June early July....

This was confirmed by a technical manager from Afrihost

Why am I not surprised...
So let me guess... so the "NEW EXCITING SERVICES" coming as per Gian, is the same old CRAP, sans the German spywall... they will probably just install a new one, probably made in China and this time, we can forget about even going near a torrent.
 
We all remember the major MTN Business Gallo Manor datacentre outage in the beginning of the month, well that was just the last straw on the camel's back for them. They use a company in Germany to manage hardware/software locally in Gallo Manor and internationally for them which deals with the prioritizing of all there bandwidth, what ports to block etc etc, you get the picture. Now the system started crashing before the Gallo Manor incident, but when that happened everything went to hell... Ever since then, they have tried everything to get the system back up and running... NO proper backups in place to roll back to nothing... I was informed that they are looking to having everything back to normal only in late June early July....

This was confirmed by a technical manager from Afrihost

61691351.jpg
 
AfriGuy,

How about you guys come clean about what really happened a few months ago... leading up to this problem?

Post from NewsHost

Ok well, I've got the 40Mb/s VDSL Businesss Uncapped with them... Was running smoothly doing around 7-8TB a month since December last year... Since beginning April I started getting **** speeds... Phoned them, and had this BS story spun to me from the techie that it’s a Telkom problem, had Telkom investigate for weeks on end at the exchange everything (thanks to a friend of mine)... Could find nothing wrong... 100m away from the MSAN.

So I started using my two test accounts with MWEB and Web Africa... Bam running at max speeds no time outs nothing. Took screen shots of everything mailed it to them. Had a supervisor phoning telling me it’s a problem with MTN... Yet another bull**** story... He told me they would put me into another IP range block... Which helped jack ****...

On Monday I decided to pop into their offices in Rivonia drive and requested to speak to a technical manager regarding my issue. Well, well, well the answer to my complaint was a bit shocking...

We all remember the major MTN Business Gallo Manor datacentre outage in the beginning of the month, well that was just the last straw on the camel's back for them. They use a company in Germany to manage hardware/software locally in Gallo Manor and internationally for them which deals with the prioritizing of all there bandwidth, what ports to block etc etc, you get the picture. Now the system started crashing before the Gallo Manor incident, but when that happened everything went to hell... Ever since then, they have tried everything to get the system back up and running... NO proper backups in place to roll back to nothing... I was informed that they are looking to having everything back to normal only in late June early July....

This was confirmed by a technical manager from Afrihost

Hi lordnokon

I'm not sure who confirmed this, but I'll happily set the record straight.
While yes, there have been poor throughput issue on accounts, it'll be investigated on a case by case basis. I'm sorry if yours wasn't given the proper attention, I'll gladly escalate an open query for you.

Around the outage at Galo Manor, this affected hosting and hosting only. Our IPC does not terminate nor breakout from Gallo Manor - it had absolutely zero affected on DSL. So I'm not sure how this got lumped into a DSL issue and somehow shaping? Regardless, hosting and connectivity environments are very different and have absolutely nothing to do with each other.
 
Hi lordnokon

I'm not sure who confirmed this, but I'll happily set the record straight.
While yes, there have been poor throughput issue on accounts, it'll be investigated on a case by case basis. I'm sorry if yours wasn't given the proper attention, I'll gladly escalate an open query for you.

Around the outage at Galo Manor, this affected hosting and hosting only. Our IPC does not terminate nor breakout from Gallo Manor - it had absolutely zero affected on DSL. So I'm not sure how this got lumped into a DSL issue and somehow shaping? Regardless, hosting and connectivity environments are very different and have absolutely nothing to do with each other.

I'll be happy to supply you with the Managers details who I dealt with?

I had a 40Mb/s Business Uncapped account with Afrihost, but only got the speeds of a 4Mb/s account.

So do you deny having a external company in Germany controlling your bandwidth, port prioritizing etc etc? And that there was a stuff-up with this and still is?
 
Well, this is like the 4th time in 4 years, I am so glad I left their sorry a s s behind at the end of 2013 with their then "massive adsl issues", never looked back.
 
I'll be happy to supply you with the Managers details who I dealt with?

I had a 40Mb/s Business Uncapped account with Afrihost, but only got the speeds of a 4Mb/s account.

So do you deny having a external company in Germany controlling your bandwidth, port prioritizing etc etc? And that there was a stuff-up with this and still is?

You're more than welcome to PM me the details.
Around your slow speeds, I'm not sure what's happened - I need to see the account history for context. Shaping is managed locally but the hardware/ software itself is international.

Saying that, shaping itself is not the cause of the issue. The cause of the slow downs recently have been an increase in demand, and while I"m the first to admit that we can manage shaping better (which is something that will happen when the POCs are launched, tested and implemented) it's not a cause of problems. On top of which, it most certainly has absolutely nothing to do with a hosting environment outage in Gallo Manor.

Hope that clears things up.
 
I used to have an Afrihost capped account that was used only for downloading of drivers for customers or MS updates at the end of every month and each month there was a issue with missing bandwidth or else when buying a topup to update the different flavours of operating systems then there is either no bandwidth available to use or there is a problem with the topup not taking place after having paid for it. Switched to Web Africa for the second account and all the problems went away, at the moment the driver downloads from the Asus website is a bit slow, it sitting at 7.1 Mbps where as it normally sits about 8.5 - 8.9 Mbps
 
You're more than welcome to PM me the details.
Around your slow speeds, I'm not sure what's happened - I need to see the account history for context. Shaping is managed locally but the hardware/ software itself is international.

Saying that, shaping itself is not the cause of the issue. The cause of the slow downs recently have been an increase in demand, and while I"m the first to admit that we can manage shaping better (which is something that will happen when the POCs are launched, tested and implemented) it's not a cause of problems. On top of which, it most certainly has absolutely nothing to do with a hosting environment outage in Gallo Manor.

Hope that clears things up.

Afrihost does not use any of their excess international hosting for their DSL network?
 
I'm not sure I follow?
Are you asking if we use international hosting for our DSL network?

Excess international hosting capacity breakout*

so they run completely separate from each other, the DSL and hosting network?
 
I don't entirely buy the explanation about running out of IPC capacity. While I'm sure that's a major factor, the last time this happened, everything was running fine until suddenly at 3pm on a Sunday everything ground to a halt for everyone. It seems a little unlikely that all of a sudden at 3pm on a Sunday afternoon demand would simultaneously increase across thousands of users to the point that the network instantly degraded. I'm no expert on large scale high bandwidth network provisioning, but I would have thought that logically you would experience a slow deterioration, not a sudden precipitous decline. In that case it was ultimately blamed during the following week on a major iOS version update. Except, here's the thing: the iOS update only became available on the Tuesday evening after that weekend, by which time the Afrihost support forums had already been swamped with complaints for two days. The way random protocols stopped working completely (e.g. in our case MySQL connectivity) while others continued to work fine, and the manner in which support would only handle individual accounts on a complaint by complaint basis where they said they would investigate the settings for that individual account, made it seem more like something was borked on the QoS or "optimisation" software level, and in fact there was eventually an admission that there was an issue with faulty software on the forums. I don't care to delve through thousands of pages of old Afrihost support forums, but it's all there.
 
I'll be happy to supply you with the Managers details who I dealt with?

I had a 40Mb/s Business Uncapped account with Afrihost, but only got the speeds of a 4Mb/s account.

So do you deny having a external company in Germany controlling your bandwidth, port prioritizing etc etc? And that there was a stuff-up with this and still is?

You're more than welcome to PM me the details.
Around your slow speeds, I'm not sure what's happened - I need to see the account history for context. Shaping is managed locally but the hardware/ software itself is international.

Saying that, shaping itself is not the cause of the issue. The cause of the slow downs recently have been an increase in demand, and while I"m the first to admit that we can manage shaping better (which is something that will happen when the POCs are launched, tested and implemented) it's not a cause of problems. On top of which, it most certainly has absolutely nothing to do with a hosting environment outage in Gallo Manor.

Hope that clears things up.

Afrihost.jpg

Yes, thanks. That does clear things up...
 
Ok, so this is why twitch has been unusable the last couple of days.
 
I don't entirely buy the explanation about running out of IPC capacity. While I'm sure that's a major factor, the last time this happened, everything was running fine until suddenly at 3pm on a Sunday everything ground to a halt for everyone.

It seems a little unlikely that all of a sudden at 3pm on a Sunday afternoon demand would simultaneously increase across thousands of users to the point that the network instantly degraded. I'm no expert on large scale high bandwidth network provisioning, but I would have thought that logically you would experience a slow deterioration, not a sudden precipitous decline.

In that case it was ultimately blamed during the following week on a major iOS version update. Except, here's the thing: the iOS update only became available on the Tuesday evening after that weekend, by which time the Afrihost support forums had already been swamped with complaints for two days.

The way random protocols stopped working completely (e.g. in our case MySQL connectivity) while others continued to work fine, and the manner in which support would only handle individual accounts on a complaint by complaint basis where they said they would investigate the settings for that individual account, made it seem more like something was borked on the QoS or "optimisation" software level, and in fact there was eventually an admission that there was an issue with faulty software on the forums.

I don't care to delve through thousands of pages of old Afrihost support forums, but it's all there.

Demand has grown rapidly, but no, it hasn't been overnight :)
We've had multiple IPC upgrades over the years but this didn't seem to be enough, this is why we're moving onto our next option.
 
Top
Sign up to the MyBroadband newsletter
X