SEACOM's SMW4 cable woes continue

Ok, enough you two - leave the ego aside for a moment please...

Ownership is not that important, the issue is cost effective bandwidth for the average user:

TRL Tech (Cybersmart) @ 9.50 per GB (50GB bundle = R 475.00) from Seacom is unreliable and the next best offering from the more reliable Eassy is provided by FNB Connect (MTN) @ 12.38 per GB (50GB bundle = R 619.00 )

Both unshaped, unlike WA / MWEB - shaped = oversold service

Alternatives?
 
Last edited:
What has mweb to do with your statement?

Glad I am still with mweb as I have no problems.

Whenever seacom was down i had no internet with mweb. the manner with which they treat their customers i won't be surprised when they start out doing telkom on hellopeter. glad you don't have any problems with them. when life gives you lemons.....
 
Nothing like 698ms ping to uk with IS. WTF? Not to mention tons of packet loss! I wish they never build seacom!
 
Whenever seacom was down i had no internet with mweb. the manner with which they treat their customers i won't be surprised when they start out doing telkom on hellopeter. glad you don't have any problems with them. when life gives you lemons.....

you probably were with them when they only had seacom bandwidth, with sat-3 i did not know seacom was down until read it here :)
 
Also been having issues here in Mozambique (Vilankulo) over the last week, definately have the long away around.

$ traceroute mybroadband.co.za
traceroute to mybroadband.co.za (41.203.21.137), 30 hops max, 60 byte packets
1 192.168.1.1 (192.168.1.1) 1.112 ms 1.726 ms 2.487 ms
2 cliadsl1-232-191.tdm.co.mz.232.138.41.in-addr.arpa (41.138.232.1) 12.639 ms 15.507 ms 24.711 ms
3 10.1.8.1 (10.1.8.1) 11.400 ms 13.768 ms *
4 * * *
5 * * *
6 * pos0-2-2.djibouti1.dji.seabone.net (213.144.175.13) 33.298 ms 34.487 ms
7 xe-8-0-0.franco31.fra.seabone.net (195.22.211.213) 159.146 ms 151.831 ms 154.043 ms
8 global-crossing.franco31.fra.seabone.net (89.221.34.50) 214.039 ms global-crossing.franco31.fra.seabone.net (89.221.34.110) 217.235 ms global-crossing.franco31.fra.seabone.net (89.221.34.50) 208.230 ms
9 MTN-DUBAI-LIMITED.TenGigabitEthernet1-4.ar8.AMS2.gblx.net (208.48.1.110) 148.439 ms 151.084 ms 149.912 ms
10 am-cr-1.nl--am-tpr-1.nl-a.mtn.net (209.212.111.140) 170.303 ms 171.216 ms 188.010 ms
11 am-tpr-1.nl--am-cr-1.nl-b.mtnns.net (209.212.111.249) 156.725 ms rb-cr-2.za--am-cr-1.nl-a.za.mtnns.net (196.44.31.139) 162.408 ms am-tpr-1.nl--am-cr-1.nl-b.mtnns.net (209.212.111.249) 159.579 ms
12 196.44.31.99 (196.44.31.99) 150.452 ms 167.786 ms tb-cr-2--lt-cr-1.uk-a.uk.mtn.net (196.44.31.212) 153.172 ms
13 ge11-0-0.gw2.jnb6.za.mtnbusiness.net (196.30.1.22) 152.234 ms jh-cr-2.za--jh-cr-1.za-a.mtnns.net (196.44.31.255) 162.361 ms tb-cr-2.za--tb-cr-1.za-a.mtnns.net (196.44.31.5) 151.035 ms
14 vlan9.hr3.jnb6.za.mtnbusiness.net (196.30.156.136) 149.852 ms jp-ad-1.za--jp-cr-1.za-a.mtnns.net (196.44.31.13) 149.682 ms 196.44.31.99 (196.44.31.99) 162.146 ms
15 196.30.213.108 (196.30.213.108) 147.317 ms jh-cr-2.za--jh-cr-1.za-a.mtnns.net (196.44.31.255) 152.068 ms 153.188 ms
16 41.72.136.53 (41.72.136.53) 150.624 ms vlan9.hr3.jnb6.za.mtnbusiness.net (196.30.156.136) 161.023 ms 41.72.136.53 (41.72.136.53) 147.734 ms
17 mybroadband.co.za (41.203.21.137) 149.100 ms 196.30.1.21 (196.30.1.21) 148.138 ms mybroadband.co.za (41.203.21.137) 149.235 ms

http://manypossibilities.net/african-undersea-cables/
 
they say here:
It is understood that since today's cable break happened outside the Egyptian curfew, the repair time should be quicker and hence the downtime won't be as long as it was on Thursday.

& its still down!! this is after just on 6hrs & on thursday last week it was only 4 hours & thursday this week i barely felt it..

when are people going to get their fingers out their asses & actuallly do something!!!
 
Been reading through this thread, and seeing many people asking about EASSY, SAT3 and other systems, so I figured I'd throw some comments into the mix.

Firstly, those who comment about cost are correct, redundancy costs money, a lot of it. For true redundancy, you need twice your normal running capacity, spread across two systems. This means that you have 50% of your bandwidth sitting idle at any one time. If you don't have that, you don't have proper redundancy, you have a service that will seriously degrade on the failure of one of the circuits. This doesn't come free, someone has to pay for it.

Secondly, if we compare the costs of IRU's (Indefeasible rights of use) on SEACOM vs the costs on EASSY or SAT-3, SEACOM has so far been cheaper. To take more capacity on either of the others, and to have it sit idle in the case of a break, is a cost that has to be passed to the consumers. Unfortunately its very often a case of consumers prefer a cheap service than a reliable service. People move ISP's because of cost, and very few of them ask "What's your backup situation" when they make their choices, they go "How much does this cost me". So the guys who have invested in the redundancy end up with the customers walking away, sad, but true. Strangely its those same customers that cry foul when things break on the cheapest service they could buy.

Thirdly, many ISP's I believe are waiting for WACS, and initial indications are that the costs involved in WACS are substantially lower than other current cables, and its West Coast based (the only other West Coast cable available right now is SAT-3, and the prices are insane. If you buy on EASSY you are STILL left with an East Coast problem that could affect all cables on the East, if you hold out for WACS, you solve this problem). Also keep in mind, these circuits aren't bought on a month by month basis, they are either long term contracts (think a year, 3 years, 5 years, or in many cases life of cable IRUs), so why invest in another east coast cable and lock yourself in when you know WACS is coming, its just about patience.

Forth, the EASSY system does not go all the way to Europe, it joins other cables further up the coast, and again, there are back to back agreements to carry them northwards. It's easy to boast 100% uptime on YOUR cable system when you're not dealing with the troubles involved the moment you have to go through "problem" areas. (Think Suez canal, think Egypt, think the med, all of which are prone to problems of anchors, breaks and other issues.) SMW3 and 4 are known for issues going through these areas.

Look, I don't like the fact that SEACOM has been down as much as it is, and I'm *FAR* from happy with the situation in Egypt at the moment (Where the hell is the terrestrial redundancy on SMW, its land based problems, its not like you have to lay another undersea cable), but at the same time I have to point out, while we whine and cry about such outages, look at your uncapped accounts today, look at what you pay your ISP's, and ask yourself, what made this possible? It was largely SEACOM, lets not forget that.

If you want fully redundant bandwidth that is on both SAT-3 and SEACOM, and hell, add EASSY into the mix, you can have it, just don't expect it at the prices you are paying for something without it, because SOMEONE has to pay for that redundancy, and when you're looking at falling margins on DSL products due to the current market price wars, the insane costs of IPC bandwidth (thanks Telkom), and the rising costs of simple things like power to drive equipment (you'd be shocked at how much this can influence costs), the ISP's can afford to do it without charging you, the customer, a lot more money.

Let's get real for a second here...
 
Nice post Andrew. Couple of questions:

1) Why is the east coast so much flakier than the west coast? Surely both have heavy traffic w/ anchors etc?

2) With additional cables, the redundancy problem should be greatly reduced right? i.e. Same amount of rands split over 3x cables should mean 1/3 degradation instead of 1/2, right?

3) Why is WACS expected to be so much cheaper than EASSY?
 
Nice post Andrew. Couple of questions:

1) Why is the east coast so much flakier than the west coast? Surely both have heavy traffic w/ anchors etc?

2) With additional cables, the redundancy problem should be greatly reduced right? i.e. Same amount of rands split over 3x cables should mean 1/3 degradation instead of 1/2, right?

3) Why is WACS expected to be so much cheaper than EASSY?

The primary reason for the West Coast being more stable than the East from what I know, and I state that I am no expert in this, is primarily because of the issues around the far north of the system. To get to Europe you either have to pass through the Suez canal or via terrestrial links to bypass that. Thats a *HUGE* shipping area, and its also not terribly deep. Go via the west and you don't have nearly the same level of problem. Passing up the east side also leaves you at this particular point in time with a number of issues when you do hit an issue. Look at what happens if you happen to have a cable break at the moment while passing Somalia, to repair it, you better have some gun boats with you or your repair ship is gonna get hijacked by a bunch of crazies. That takes money and time to organize.

With regards to more cable systems, in theory yes, the more cables an ISP has capacity on, the less the degradation will be when they lose a portion of the capacity, but there WILL still be degradation. Also, keep in mind that with the costs of the cables being so far from equal to each other, ISP's, just like their customers, will look for the most economical options and potentially buy more capacity on a cheaper cable and less on the more expensive ones, so there is no guarantee of an equal three way split.
This is further complicated by the fact that you get *VAST* discounts if you buy capacity in volume, so it is *FAR* more expensive to buy say, 3 x STM-16 (2.5gig) links, than 1 x STM-64 (10gig) link, purely because of the volume purchasing discounts. As a result of this, to keep the costs down, you need to buy volume, which doesn't leave the option to split between cables.

With regards to your third question, again, I believe this comes down to sizing. If you build a cable system, you put a number of fiber pairs in it. Each fiber pair you add to the cable makes the cost per megabit / second cheaper. The costs are also drastically changed by the number of landing stations you have to build and how many break out points. SEACOM itself if I recall is a 2 pair cable, WACS is a 4 pair I believe. The difference in bandwidth terms on these cables is vast, with SEACOM being either a 1.2 or a 1.4 terabit system I believe (again, open to correction), WACS is three or four times that size, so, again, volume build = lower price.

There are so many complex issues involved in the economics of bandwidth provision and the consumer needs to be cognizant of this, if you're buying a service, and its the cheapest available, you're *GOING* to have problems sooner or later, so the consumer needs to do some homework into what he is buying. What are the contention ratios? What are the backup plans? What is the reliability? Once that homework is done, the consumer can ask himself if he's willing to accept the risks, or pay the premium that comes with buying something of higher quality.
 
No problems on MWEB - speedtest.net to London as usual.

Anyone on MWEB still having hassles?
 
There are so many complex issues involved in the economics of bandwidth provision and the consumer needs to be cognizant of this, if you're buying a service, and its the cheapest available, you're *GOING* to have problems sooner or later, so the consumer needs to do some homework into what he is buying. What are the contention ratios? What are the backup plans? What is the reliability? Once that homework is done, the consumer can ask himself if he's willing to accept the risks, or pay the premium that comes with buying something of higher quality.

Accepted, however there are quite a few local ISP's that I can list that are dishonest and ambiguous in their dealings with the consumer. Have you ever tried obtaining contention ratios from any local ISP?

Best advice I can give to the consumer is:

Stay away from ISP's that utilise shaping and or utilise IS' backbone.
 
Last edited:
Thanks for the comprehensive answer. :)

Makes sense w/ east vs west.

Volume discount...didn't think of that. That suggests to me that the ISPs should pool there resources. i.e. buy 3x STM-64s and then split that 3 way. That way they get the discounts and the redundancy. I'm guessing cooperation between ISPs is easier said than done though...

With regards to your third question, again, I believe this comes down to sizing.
I hope you're right about WACS being cheaper, but I'm kinda skeptical about the size argument mainly because of the % of fiber unlit. Sure you'd need more endpoint equipment which costs, but until its 100% lit I don't think the total b/w capacity should affect pricing. Maybe I'm misunderstanding the unlit thing though...
 
they say here:
It is understood that since today's cable break happened outside the Egyptian curfew, the repair time should be quicker and hence the downtime won't be as long as it was on Thursday.

& its still down!! this is after just on 6hrs & on thursday last week it was only 4 hours & thursday this week i barely felt it..

when are people going to get their fingers out their asses & actuallly do something!!!

Thursday was down from 23h30 the night before until 8h10 when the redundancy came up. Full cable repair was 4 hours later.
 
ye me too... im on a 4mb line & im barely able to get even 70kb/s now..

check my speedtest results

well this is the new ping test..this is to jhb, dunno how right it is: http://www.pingtest.net/result/37281828.png

and this one is to london uk: too much packet loss error :P

the speedtest page wont load
 
Top
Sign up to the MyBroadband newsletter
X