Last Friday about 11:00 my Internet connection failed. I contacted the ISP and asked why the DSL was offline. They answered that there is a problem, they do not know what caused it and hopefully it will be on again by the end of the day.
The loss of connectivity during a working day is detrimental for two reasons: customers lose valuable work time and customers tend to start looking at other ISPs.
I asked the ISP first level support person was what the problem was. They did not get told what the problem was and therefore could not say. This lack of information only adds to a customers frustration.
The regulator does not force the ISP to publish uptime ratios. Therefore the customer does not really have any way to judge the ISP performance, especially if connection reliability is the most critical issue to a customer. Price appears to be very similar between ISPs so this is often not an important aspect of the customers choice in ISPs.
In this situation the DSL product is a carriers offering being resold by a smaller ISP. The DSL connectivity failure appeared to happen upstream of the DSLAM, authentication server and first gateway on the ISPs network. It could therefore have been a problem caused by the carrier. Likely? very likely.
The other important issue highlighted by this network failure is the lack of redundancy in the DSL network. If the network design was adequate and sufficient traffic pathways had been in place there would have been no outage of significance, a few minutes possibly.
On Saturday the ISP technical support starts at 11:00 and the network was still not up at 10:50, nearly 24 hours since the network outage started. I settled in with a cup of tea ready to call at 11:00 and like a miracle the network started functioning again at 10:55, just in time to stop a flood of support calls. This is a touch too obvious to be an accident of timing.
Was there an outage at all or just simply a network upgrade that meant that customers would have to go offline whilst it was done? Surely the carrier would not do this? The carrier would if there is only one set of hardware and one traffic path to the ISP.
Was this an isolated incident? I hear from people that I talk to that this occurrence happens all the time across most ISPs in my country. It is time for the regulator to force ISPs to publish their uptime ratios so that customers can see this important reliability data.
regards, Mark