[Gllug] [OT]fire at redbus?

[Tethys]

Liam Delahunty writes:

>There might have been a fire, I have a daily log of the CPU temp sent to 
>me in the morning. Just having a look now, One can see the temp rising 
>see below, the server normally runs at about 36 during the day, and 
>heats up at 4 am with all the crons/log rotations.

Of course, a temperature increase could be explained by aircon failure
just as easily as by a fire...

Is it just me, or is anyone else concerned about power failure in their
hosting centre? I mean, our hosted machines have two independent power
feeds, with near instantaneous failover between the two. But even if
those fail, the whole datacentre has UPS backup via a massive battery
array in the basement, and two diesel generators to back those up.
The prospect of that level of redundancy not being enough to keep the
juice flowing is somewhat... disturbing!

Interestingly, Redbus say:

	All our facilities are built to at least an N+1 standard; put simply,
	this means that if two generators are required to provide non-stop
	power, then there'll be three installed. So rather than guaranteeing
	99.99% availability, the guarantee is 100%.

Failing to meet a 100% uptime SLA? Could prove expensive for them...

Tet
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug