SEACOM and Cybersmart set record straight
Well known Internet Service Provider (ISP) Cybersmart has been testing SEACOM with the aim of integrating the more affordable international bandwidth into its network and drive down prices. High latency and stability issues on the cable have however seen the ISP put their SEACOM integration on ice.
A recent email from Cybersmart to its subscribers pointed out that the ISP experienced “four long outages” over the last few weeks, and that the “195ms contractual latency between Cape Town and London is not even close to being achieved”.
“Whatever the reason for the outages and no matter who is responsible, the service at the moment is certainly not viable and in good conscience we cannot put our customers onto it no matter what the cost savings might be. The unfortunate fact of the matter is that the marketing hype around this new cable does not meet reality,” the Cybersmart email read.
“We will be moving off Seacom back onto SAT-3 over the weekend and will remain on SAT-3 until such time as the contractual obligations in our contract are met and the network between Cape Town and Mtunzini becomes stable.”
Cybersmart MD responds
This email sparked reports that the ISP has ‘dumped’ SEACOM, something which Cybersmart CEO Laurie Failkov said is certainly not true. “Firstly I want to put on record we made no public announcement, nor did we say we went back to buying from Telkom. We sent out a network notification saying that we were going to have two maintenance windows. As with all our network notifications we give detailed reasons as to why the maintenance windows are required,” said Fialkov.
Speaking about a recent media report about the development, Fialkov said that someone was kind enough to post this network notification on the MyBroadband public forums “and everything else attributed to Cybersmart is nothing short of fanciful elaboration and in some cases fabrication.”
“Some of Arthur Goldstucks comments regarding what Cybersmart said makes it blatantly clear that he did not read the notification and it is quite disturbing that he found it appropriate to comment on it,” said Fialkov.
“The reason we did not buy from Seacom directly was twofold. We knew Seacom were not terminating in Cape Town and we wanted to have a competitive advantage by being the only provider to have ‘cheap Seacom bandwidth in Cape Town’,” Fialkov explained. Secondly, and more importantly, we wanted a single contract which covered connectivity from our data center to Telehouse in London, including IP transit.”
“We knew that if there was a problem and there was even a remote possibility that it could not be attributed to the actual undersea cable, Seacom would throw up their hands and say ‘it is not us’. The comments attributed to Brian Herlihy have shown that our strategy of an end-to-end contract was a sound business decision. Buying an international contract without an end to end service guarantee is commercial suicide. We do not have customers at the landing station,” says Fialkov.
Fialkov added that they are no longer happy to sit back and accept criticism that they are “SEACOM bashing”, and presented the “following facts”:
- The backhaul from Mtunzini (SEACOM endpoint ) to Midrand is provided by Neotel
- The backhaul from Mtunzini to our Cape Town datacenter was also provided by Neotel
- The backhaul from Melkbos (SAT-3) datacenter is also provided by Neotel and the SAT-3 is not provided by Telkom
- The round trip delay router to router (Cape Town, Mtunzini, SEACOM, London) fluctuates between 260 to 320 ms
- The round trip delay router to router (Cape Town, Melkbos, SAT-3, London ) fluctuates between 148 and 152 ms
- The link Cape Town – Mtunzini – London was recording 2 CRC errors every hour on our SDH interface
- The link Cape Town, Melkbos, SAT-3, London has yet to record any CRC errorsSEACOM claim everything is 100% however they have yet to show a router in either Midrand or Mtunzini (and they have other customers there) with no CRC errors and we have asked for this information to substantiate their claim that it is working 100% to Midrand.
- I can’t recall there ever being a complete outage on SAT-3 or the backhaul
- In the first 3 months of trading there has been a 2 hour plus total outage on the SEACOM link, whether the fault was a result of SEACOM or their backhaul partner Neotel is not relevant to the customer
- Neither SAT-3 nor SEACOM terminate in our datacenter – both require backhaul links from their respective landing stations. It is common practice to say that there is a failure on the SAT-3 link or a failure on the SEACOM link irrespective of whether the cause of the failure is as a result of the backhaul or the undersea cable. There is no preferred supplier for SAT-3 backhaul however SEACOM have gone on record on numerous occasions stating that Neotel is doing their backhaul
- If the situation was reversed and we notified our customers as to how fantastic SEACOM is and that we are moving all our capacity onto it (a situation we were hoping would occur), there would not be any mention of “Cybersmart buying from a third party ” or SEACOM distancing itself from its partner Neotel
“Having said all of the above the SEACOM project is an admirable feat of engineering and we believe in its ultimate success wholeheartedly. If we did not believe in it we would not have invested in it and reduced our capacity with Telkom,” said Fialkov.
“Not once did our ‘third party supplier’ ever try to avert blame and they (not Telkom) offered us SAT-3 capacity as an interim solution until they could fulfill the contractual end-to-end obligations and ensure the technical specifications in the contract could be met. We have not ‘dumped Seacom’ as some articles claim, [but are] in fact more than happy to offer whatever assistance is necessary in order get Seacom to a position where it is comparable to SAT-3, not just to the landing station but every end user throughout the country,” Fialkov concluded.
SEACOM responds to instability claims
It is however not only Cybersmart which raised concerns about the stability of SEACOM bandwidth, as explained in a recent article: http://mybroadband.co.za/news/Telecoms/10098.html. SEACOM however said that the downtime is not related to the main cable system, but rather to the terrestrial backhaul network.
“It is important to note that the SEACOM Cable System was not at fault and has been fully operational since inception with no failure. We need to be mindful of who is accountable for failures,” said Ramdhani. “The SEACOM Cable is but one link in the complete service that customers purchase in parts to their door steps.”
Neotel confirmed that the recent downtime was caused by network failure from its side: “There are three routes available to protect the SEACOM service, but this morning two of them were down. It automatically shifted to the second route, but there was a second break on the alternative route between Port Elizabeth and Cape Town, therefore we have now had to manually shift to the 3rd route which runs through Durban, Pietermaritzburg and Bloemfontein.”
Ramdhani added that some companies which operate on SEACOM capacity have access and national links which were not purchased as a bundle from SEACOM, which means that SEACOM cannot be held accountable for many of the instability issues which are experienced.
SEACOM CEO Brian Herlihy says that they have had several discussions with Cybersmart and better understand their situation. “Cybersmart purchased a product from TATA/Neotel which serviced them from Cape Town – a product that SEACOM does not yet offer. This product has had several outages since they went live,” says Herlihy.
Herlihy explains that none of the outages referred to by Cybersmart are with the SEACOM cable, and that they had 100% up time since the launch. “[Cybersmart] stated in our meeting with them that this is not a SEACOM issue but an access to SEACOM issue. The TATA/Neotel product which they bought has been appropriately dealt with by TATA/Neotel, which means TATA/Neotel have put them on a SAT-3 route until they can stabilize the national network issue,” said Herlihy.
The SEACOM CEO further questioned the accuracy of recent media reports about SEACOM’s stability. “I do think the press has been unfair in this process (or not diligent in this matter); for example, when a service goes down from Telkom we don’t assume that SAT-3 has gone down (even though it has gone down several times in the last six months with each occurrence being an appropriate report in and of itself) but in this instance when a national service goes down it is being reported as a problem with SEACOM.”
“There seems to be a double standard and I am surprised that more investigation isn’t being done to see if SEACOM is down or not. Ultimately I do believe that the national networks will increase in quality of service and that the total package will become better but, today, I think it is important for everyone to understand what part of the total package is truly being effected.”
Herlihy said that they had a number of potential customers calling SEACOM with very nervous reports on comments from the Cybersmart report on MyBroadband – both about latency and down time. “We have been able to demonstrate to these customers that both of the concerns are unfounded. I think during the launch and the SEACOM experience all reporters understand what SEACOM’s real latency times are. We have consistently showed our RTT to London at or less than 180 ms, but this one incorrect posting has created a lot of misinformation with the institutional players.”
SEACOM & Cybersmart discussion
Loading ...