SEACOM downtime after local backhaul problems

Afrihost seems to be working without the proxy now... but I am still considering moving as this is likely to happen again as history is proving!
 
ToxicWazte said:
Does nobody else here but me get that this is Neotel's f-up, and not SEACOM's?????????

SEACOM is a submarine cable, i.e. it is at the bottom of the ocean. Neotel are the idiots that SEACOM has unfortunately decided to partner with landing the cable and for terrestrial backhaul from Mtunzini to JHB. This part of the network is NOT part of SEACOM - this is part of Neotel's network. SEACOM ends in Mtunizini, on the beach.

That said, SEACOM should really consider scrapping their partnership with Neotel, and going with Telkom, or DFA instead. It was a really poor choice.
Click to expand...

Correct, since TENET who partnered with DFA, have not had any downtime today, as per this traffic graph:
http://monitor.net.tenet.ac.za/cacti/graph_view.php?action=tree&tree_id=6&leaf_id=874
 
I'm unsure as to why, in this day and age, a backup fiber route has a MANUAL changeover? WTF? Go Neotel!

My IS based accounts were up during the April outage and today, maybe we get lucky packet access to the Satellites?
 
ToxicWazte said:
Does nobody else here but me get that this is Neotel's f-up, and not SEACOM's?????????



SEACOM is a submarine cable, i.e. it is at the bottom of the ocean. Neotel are the idiots that SEACOM has unfortunately decided to partner with landing the cable and for terrestrial backhaul from Mtunzini to JHB. This part of the network is NOT part of SEACOM - this is part of Neotel's network. SEACOM ends in Mtunizini, on the beach.

That said, SEACOM should really consider scrapping their partnership with Neotel, and going with Telkom, or DFA instead. It was a really poor choice.

Also, I don't understand why IS do not have redundancy on SAT3 and via satellite. It's just plain stupid.

The companies that are the most to blame for this issue is firstly Neotel, and secondly IS. Most of the other ISPs are just resellers of IS bandwidth.$\
Click to expand...

I was thinking the same thing. Words can't express how much I dislike Neotel. The irony is that I have had dealings with Telkom over the past few weeks and they have been a pleasure to deal with. What is going on here? Neotel were supposed to be great but have actually turned out to be a little rubbish...
 
pkid said:
I was thinking the same thing. Words can't express how much I dislike Neotel. The irony is that I have had dealings with Telkom over the past few weeks and they have been a pleasure to deal with. What is going on here? Neotel were supposed to be great but have actually turned out to be a little rubbish...
Click to expand...

People don't plan to fail. They simply fail to plan, except Neotel, who just don't plan and fail
 
Biscuit1018 said:
This has affected Afrihost significantly
Web Africa is fine

How is MWeb guys?
Click to expand...

The last outage affecting my AfriHost line was extremely frustrating, especially as it was near month end... and now this? Eish! :(

If this is going to become a regular feature of Seacom, I will be switching to another ISP. Sorry to AfriHost, but I must have my internet.
 
JustinHarrison said:
You would think after the last outage Afrihost would have made some preparations in case of another outage... this is insane!!!!!!!!!!! - Which ISP does not use Seacom, because I am ready to move!
Click to expand...

This was one of my recent threads, how many people will jump ship and move back to isp's on sat3?
 
There is also HostworX.co.za who offers DSL Uncapped Accounts, we use them and we get full line speed out of their accounts, unlike the issues seen with other uncapped accounts, ie afrihost and screamer, their accounts were plain just sloooooow, my IS account is also a thing of the past :D

So if you looking for an altenative provider they have my vote. (BTW my account with them was not affected bu the seacom and IS outages, so they definately not using them):erm:
 
Hmm our Afrihost router at work is not connecting - can't reach it at all. Maybe it needs a restart.
 
Afrihost Speed Test: (4mb capped)

New York sux
London sux worse than a 384 connection

Midrand ??
JHB
CPT like a 1mb connection ...

Big WTF ?!?
 
Last edited:
Gordon_R said:
Correct, since TENET who partnered with DFA, have not had any downtime today, as per this traffic graph:
http://monitor.net.tenet.ac.za/cacti/graph_view.php?action=tree&tree_id=6&leaf_id=874
Click to expand...

Just a note and a coupla comments. Firstly, what Gordon says is true, these problems have not affected TENET. However, it also needs to be said that TENET has had its fair share of outages on its international bandwidth due to fiber breaks with idiots attempting to steal the fiber thinking its copper. We've just implemented a redundant path to the landing station night before last actually to solve this as well.

With that said, SEACOM itself has been remarkably stable. Apart from the SMW4 outage (which was completely out of their control), there have been two or three other very brief incidents that resulted from hardware issues in mumbai (power feeder issues), but these outages have always been very timeously fixed and have never lasted more than a couple of minutes.

With regards to other ISPs and lack of redundancy, in their defense and this needs to be very clearly stated, the reason ISPs don't have enough capacity to carry everything via a redundant path is because of cost. If your read these forumns, everyone on here cries for lower pricing all day every day, except they forget, lower pricing comes at a tradeoff against things like redundancy and additional capacity.

In order to run 100% in the event of a cable failure on one cable, you have to permenantly keep in reserve the same amount of capacity on another cable system, this means you are effectively overprovisioning your network at a 2:1 ratio. This costs money, that some how has to be recovered, and we're not talking about minor money either. (Current Telkom cost on SAT3 if I recall is around R550k per 155mbit STM-1 (140mbit useable)), This means that if you are normally running at 1.5 gigabit, you'd need to be burning 5+ million a month in capacity that is effectively sitting idle and earning you no revenue, and while that can be done, someone has to pay for it, and thats going to drive up costs.

Will EASSY help in a few months time? I have my doubts, since EASSY is a consortium of a ton of African telcos all looking for their piece of the pie, and while I'd love to believe that EASSY is going to come in nice and cheap, I have a strong feeling that we'll see it as just yet another very expensive option.

The true redundancy will arrive with the coming of WACS. Not only does WACS run on a different physical route from SEACOM/EASSY, it also has huge economies of scale involved considering the size of the cable.

It's unfortunate that these problems are occuring, but guys, change takes time, and I might point out that 2 years ago, what we see today in terms of cost was virtually unheard of, give it time, these things will come right.

Btw, as a matter of interest, to demonstrate just how much SEACOM has changed the landscape of things already, on March 15th 2008 the total bandwidth TENET was delivering to its institutions totalled less than 180mbit/second, today, we are actively delivering more than 1.5gigabit/second to our institutions and growing every day, with a 300% growth in active traffic in the last 6 months alone.

Yes, its frustrating that the outages occur, but bear with it, these things will be fixed in time as more competing cables come up and as the reliance on SAT-3 as the only realistic backup method is decreased if not eliminated.

Just my thoughts
 
AndrewAlston said:
Just a note and a coupla comments.

<<snip>>

Just my thoughts
Click to expand...

Thanks for the detailed repsonse. Much of what passes for comment on this forum is lacking in substance (facts, costs, etc).

Just a thought: If there are 3 redundant paths, then the reserve capacity drops to 3:2 (1.5:1), which would be a lot more cost effective. In this respect, the more cables the better.

I think for most users the total lack of transparency ( and 'buck-passing') during these breakdowns, is the most frustrating part. AFAIK no ISP (other than TENET and JINX) publicly posts traffic graphs.

P.S. I did notice that TENET had to switch to its secondary path yesterday (I didn't record when), but the downtime was quite short.
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X