SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Pastimes : Dream Machine ( Build your own PC )

 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: Jay Fisk who wrote (506)4/23/1998 9:14:00 PM
From: Spots  Read Replies (1) of 14778
 
Thread, a sad but true story about UPSs.

Monday this week I had THE power outage I have a UPS for
(I was about to FTP a software fix for a customer trouble
ticket, but that's incidental). Just on the moment, there
was an area power failure. Here are the sad details:

1) Although I run "regular" tests of my UPS, I haven't done
so in the last few months (I'd like to claim three months,
but probably it's six).

2) When the power failed, my monitor wasn't plugged into the
UPS. So blanksville. This can be disconcerting. I don't know
how this happened; I "always" keep my monitor on the UPS.

3) When the power failed, my workstation CPU ran for (perhaps)
a minute, then shut down. This was without the monitor load
on the UPS, of course. This is without even looking at the
server that I run off the same UPS.

4) I verified all this by plugging in a 75-watt incandescent
light directly into the UPS. Not a glimmer.

5) This is an APC 600 va Network Power Station UPS. An older
model but one which I bought about a year ago.

6) I called APC and they told me my unit had a one-year warrantee
(although the book said two years). I pulled the book on them,
and they agreed to replace it, but if I hadn't had the book ...
well. Let me not fault them too badly here; the dern thing
failed and I had a minor hassle, but they are cross shipping
a replacement (if I don't ship the old one back in 30 days
I get a credit card charge, but that's reasonable to my mind).

7) My test procedures sucked. In the first place, I should
have done them every month instead of every few months.
Batteries CAN fail and there's no quality control in the
world that can prevent that in some percentage of samples.

8) Even sadder than 7), in may last test (goodness knows
how many months ago), the UPS ran my protected systems
for 9 minutes rather than the typical 13 minutes of earlier
tests. All I said to myself was, "Hey, that's plenty"
instead of saying "Does that make sense? Have I added THAT
much equipment?" If I'd asked the second question, I
probably wouldn't have demanded a replacement at that point,
but I (hope) I wouldn't have let another four or five months
go by before the next test (or, in this case, the failure).
Then I could have (probably) documented 13 minutes, 9 minutes,
7 minutes, 5 minutes, HEY we got a problem.

9) APC has a great reputation, but have your paperwork handy.

10) The crux of the matter. It doesn't matter how good your
equipment is or how great your preparation is or anything else
about your setup if you screw up your procedures.

The only successful procedure to emerge from
this at my place is the way I keep my hardware paperwork handy.
Ugh (actually, I have to admit I was lucky about that--normally
I'm terrible at paperwork; ask my wife! BTW, you CAN'T keep your
hardware paperwork [especially UPS] on the PC! Does this seem
obvious? Well, I won't tell you how it used to be here ...).

Anyway, maybe my paperwork is great (yug), but I'd rather have
done better tests. Shoot, it only takes 13 minutes a month ...

Sadder but wiser,

Spots
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext