Comments on: Power Outage at Gnax http://blog.apisnetworks.com/2008/01/09/power-outage-at-gnax/ Your one-stop place for Apis Networks updates Fri, 05 Sep 2008 21:10:45 +0000 http://wordpress.org/?v=2.5.1 By: Apis Networks Community Updates » Data center power outage/kernel upgrade tonight http://blog.apisnetworks.com/2008/01/09/power-outage-at-gnax/#comment-115 Apis Networks Community Updates » Data center power outage/kernel upgrade tonight Wed, 21 May 2008 00:45:36 +0000 http://blog.apisnetworks.com/2008/01/09/power-outage-at-gnax-unconfirmed/#comment-115 [...] of this year, I thought — as well as everyone else — that that would be Gnax’s only blunder with their UPS system.  Gnax either cannot learn from their lessons or is mired by incompetence higher up in the [...] [...] of this year, I thought — as well as everyone else — that that would be Gnax’s only blunder with their UPS system.  Gnax either cannot learn from their lessons or is mired by incompetence higher up in the [...]

]]> By: David Kozinn http://blog.apisnetworks.com/2008/01/09/power-outage-at-gnax/#comment-23 David Kozinn Thu, 10 Jan 2008 03:54:32 +0000 http://blog.apisnetworks.com/2008/01/09/power-outage-at-gnax-unconfirmed/#comment-23 What I do find impressive is that they actually seem to know what happened. At my "day job" they've had some pretty serious issues (thankfully, nothing like this), and there tends to be a lot of confusion for a couple of days, part of which, unfortunately, seems to be caused by folks who are trying to spin things so that it doesn't sound like it's their fault. When my company moved to (their own) new data center this year, once of the things they did was go to out of band management for every single piece of gear in there. In the racks, they have dual-power (A and B feed, as GNAX described) intelligent PDUs that are both on separate out of band networks solely for management. Of course, nothing guarantees something bizarre couldn't occur there, but then again, our data center is purely for in-house use, and we've always got something like a half-dozen (or maybe more) in the shop "just in case". I am going to take the GNAX explanation and run it by a couple of people here with the question "could this possibly happen to us"? I'm not responsible for the data center itself, but I am responsible for several key infrastructure components that enable our workforce to do their jobs. As a big "customer" of our data center, I want to know if I've got anything to worry about. Thanks for posting that Matt. What I do find impressive is that they actually seem to know what happened. At my “day job” they’ve had some pretty serious issues (thankfully, nothing like this), and there tends to be a lot of confusion for a couple of days, part of which, unfortunately, seems to be caused by folks who are trying to spin things so that it doesn’t sound like it’s their fault.

When my company moved to (their own) new data center this year, once of the things they did was go to out of band management for every single piece of gear in there. In the racks, they have dual-power (A and B feed, as GNAX described) intelligent PDUs that are both on separate out of band networks solely for management. Of course, nothing guarantees something bizarre couldn’t occur there, but then again, our data center is purely for in-house use, and we’ve always got something like a half-dozen (or maybe more) in the shop “just in case”.

I am going to take the GNAX explanation and run it by a couple of people here with the question “could this possibly happen to us”? I’m not responsible for the data center itself, but I am responsible for several key infrastructure components that enable our workforce to do their jobs. As a big “customer” of our data center, I want to know if I’ve got anything to worry about.

Thanks for posting that Matt.

]]>