Why 12 hour outages can sometimes be good

12 hours ~= 3a’s worth of 99.95% SLA allowable downtime
Graph produced by Smokeping. Click for full-size.


Remember that 99.95% service-level-agreement (SLA) I was talking about last week? And that power-failure? Add a 12-hour network outage and you’ve got serious problems.
Just like that, these guys are moving into Server North’s datacenter in Toronto. How’s that for pwn? 🙂

Seems that Hydro-Telecom (Telecom-Ottawa? Trytel?) forgot, during ‘network restructuring’, to route a /26 towards the datacenter my client is currently residing in. Their suggested solution was to setup a routing protocol (like BGP) between Hydro-Telecom and the guys managing the network in the building… except I know they have neither the inclination, budget or skill-power required to do that. (No insult to the guys running the show, but have you actually setup BGP4 before?) And it’s been made clear to me that that’s not their business model. Arg.

Well, maybe not so ‘Arg’ anymore:

I’m being sent packing with a half a rack’s worth of equipment to Toronto on Wednesday night… this has to be the fastest colo agreement I’ve ever negotiated – but these guys can’t afford to mess around anymore. Chris and I have spent most of today working out the Ts&Cs and all that jazz.

Things are kicking it into high-gear, so I expect I’ll be in Toronto aplenty over the next several months. Again.

City of York – when will you adopt this small-town boy?
I think I made 13-14 trips to Toronto in 2005… so far in 2006… I’ve been once. It actually seems kinda weird… A few months ago I found myself missing Toronto. Sick? At least I have an excuse to stop by Saigon Delights in Kingston again! 🙂

Posted in Uncategorized