The Setup
A few minutes ago, my old APC UPS 875′s Replace Battery alarm fired, triggering an enterprise-wide alert and making a blaringly-loud announcement throughout the entire facility.
I raced down to see what was amiss (I’d never heard the alarm before, so I was very surprised… although Hal and Wanda didn’t seem terribly fazed by it… as if they had heard and felt it before), and located the source of the ruckus.
The Premise
Fortunately, I happened to have a ready-to-go, new “green energy” APC UPS 750 (with the master-slave socket configuration, no less!) sitting prettily at the loading dock. It was originally acquired in order to provide a steady power stream for the LCD TV, Xbox 360, and the rest of the home theater rig.
This goes back to that old adage, “You can never have enough power!” This was, of course, one of our old tekkie shout-outs from the 1980′s (argh, this kinda dates me, doesn’t it?), but it is as true now as it ever was.
However, I never seem to have a UPS on standby, ready to be hot-swapped into position whenever any of my currently deployed ones gives up the proverbial ghost. But why not? It would make sense, right? In the past(s) of my professional careers, I’ve characteristically ordered multiples to guarantee that there’s always redundancy. Nothing like 100.00000% uptime, or at least 99.99999% to make customers and contractees happy. So why skimp back at HQ?
The Action
This event certainly falls into the category of “emergency”, so I figured, “Let’s get this baby started!” At this point, I’m referring to the APC UPS 750. I check with Hal and Wanda; they both nodded YES!
So, four minutes later, the APC UPS 750 was unboxed, battery terminals connected, initial bootup completed, wired into the monitoring network, and ready to go. I had to shut down the old APC UPS 875, which unfortunately caused the immediate shut down of one of the backbone network switches as well as the Dell PowerEdge SC420, which in turn forced the shutdown of several of the main network services, like the primary DNS server, a FTP server, a few Web servers, and a few Samba shares.
Unplug, unplug, unplug– ad nauseum; plug, plug, plug– again, seemingly ad nauseum.
The new UPS is in place, power is reestablished, and the network switch and the Dell server are back in action. All is well with the universe once again.
So What’s the Next Step?
“You can never have enough power!”
Gee, it sounds like I read/heard/saw/thought that so recently… and it’s still as true now as ever. A little bit of shopping is in order; now in addition to the UPS for the home theater, it would make a lot of sense to have another UPS for emergency standby purposes.
A key concern is how much capacity AND sockets are required in an emergency UPS. For different people, this may vary, but for my purposes, the possibility of an “emergency UPS” becoming the permanent replacement UPS is extremely high. My much older area UPSes range from the 450 VA to 500 VA, while my newer area UPSes range from 500 VA to 750 VA. My stack and server UPSes tend to be between 750 VA to 950 VA. I’ve never really standardized on any given manufacturer, but that may be something to consider much more seriously in the (near) future. Historically, I’ve hovered between APC and Tripp Lite, and I’ve dabbled with Belkin and Conext.
It may make sense to have a couple of UPSes available for emergency standby, then, where two may be deployed to provide coverage for a dying UPS (so, a 1:2 substitution), and if neither is suitable as a permanent replacement, at least there’s some time available to shop for the correct replacement UPS.
