Well the irony is when we enable additional capacity it seems we are being routed via the incorrect cable system. This has only been happening the past few days. We are killing the burst and routing another path to better the latency. We are cancelling our current contract and moving our layer2 to London.
PTSD is the life of an ISP, I'd like to think honesty is the the point here.Going to give you the benefit of the doubt just because of how active and open you seem to be on the forums here but this type of thing triggers my PTSD with a certain red ISP I used to be with
Clearly not even when showing comparitive results where internal graphing doesn't match and then 3rd party customers don't match yet it's still the ISPs doing, I have sent the same graphs and so have other customers which provide skewed results on last miles?
This means we have additional capacity in the evenings, when we require additional capacity it is being routed over eassy routes and not sat3 or wacs as it should. So every 3rd packet routes via eassy from what we can see.
So our layer 2 is split into 3 portions WACS, SAT3 and then burst capacity. Burst is proving to be a latency spike which is supposed to be via a WACS/SAT3 blend but it is causing latency spikes so we are moving these services.
So the way to get capacity cheaper is to buy from large business ISPs that use capacity during the day and not in the evening. Then leverage evening capacity when we require it. It's all about delivering capacity when you need to at the most cost effective rate. This portion is now an issue so we have dropped it and routing it via our IS capacity instead whilst we replace it. We have had this capacity for nearly a year without any issues.
Jaco, all I've provided graphs to match the time spans of your own. And at the same time other customers on other last miles have different results. I don't see how Openserve are the only ones that have assisted here, when I have been open about everything we get provided and pass this onto you? The configuration between OS and an ISP is a simple one, not something that can be mixed up between operators. We deal with more than 10 operators and OS is the most difficult, most expensive and least professional we have dealt with. You are welcome to give your opinion as am I having dealt with every operator. I have no reason to lie about anything.
ISPs get capacity from carriers and decide how to manage them. We are having an issue with a particular carrier, this is very much normal. And we are adjusting the way traffic is routed to avoid said carrier whilst we change that route.
Carrier to your house is OS?
One out of 3 pings would route via an alternative path on our international routes, I will blame OS when local tests don't correlate to 2 sets of alternative graphs on local tests, And other 3rd party networks have different results at the same time. We seem to be jumping around what the issue seems to be here?
Sign up with Google