[nSLUG] Power Loss and UPSs

D G Teed donald.teed at gmail.com
Sun Nov 4 21:24:16 AST 2007

We were the site with about 2 hours of backup power time.
It turns out we didn't lose power except for some flicker,
which was likely lightening strikes.  I also heard that
our 2 hours was useless anyway because the air conditioning
loss would cause our servers to cook after about 30 minutes
or more.  At my work, we really need to talk about a
shutdown client implementation.


On 11/4/07, Jim Haliburton <jim at on-site.ns.ca> wrote:
> Hello all:
> Well it is mostly over as I write this on Sunday evening.  This afternoon
> was delightful in Metro as those around here could see.  In my office the
> power came on between 1:00pm and 2.  Am not sure as I was out.
> I am impressed that someone on the list has UPSs that will hold 50+
> servers up for 2 hours.  That is a lot of battery power.
> I am not so lucky.  On Friday I picked up 3 gallons of fresh high test
> gas for my backup generator.    Sometime after 5:00 this AM I realized the
> clock was not visible and as I bcame fully concious could hear the
> symphony of beeping coming from the server room which is almost under my
> bedroom.
> Scrambled madly to find the flashlight.  Threw on some footwear and went
> and checked the status of all servers.  All were still up and running and
> could see the world.  Quickly fueled the generator and ran the power line
> to the server room.  Once the generator was started connected up, the
> wailing of the alarms stopped.
> At this point I found the Novell server's UPS had run down as had my
> desktop system.  I just powered the UPSs off and left them.
> Some post mortem points and items about the other responses.
> 1)  I should have fueled and test run the generator on Friday.
> 2) I should have run the extension cord to the server room Friday as
> well.
> 3)  Should have made sure I knew where the flashlights were last night.
> 4)  Really must track down the UPS control software for my all my
> servers.
> You can apply these observations to your business case.
> Test the UPSs before they are needed.  I know the Novell version of the
> software I had before would allow you to under software control switch
> over to battery power and monitor the life left as the voltage dropped.
> The app ran as an NLM on the server and a Windows app on a workstation
> talked to the NLM.
> Plan ahead and decide which servers are needed to stay up and which can
> be powered down to extend the run-time of the UPSs.  I should have done
> this Saturday evening.
> Can virtualization allow you to reduce the server count in times of power
> instability?
> Can you move to a hot site easily if the power outage is accompanaied by
> infrastructure damage that will last longer term.  For example if the
> roof is blown off over your server room, where do you move to?
> From this I will refresh my planning.  Even as a small web hoster I have
> obligations to clients to keep running.  After Juan, my little generator
> ran 110+ hours without a break.  Once my ISP connection was re-
> established we served our clients without shutdown.  And without power
> from NPSI.
> It would be intersting to hear from others about their post mortem of the
> power outage.
> Now most of my UPSs are back to 90+% fully charged.  Over the next few
> weeks I will test the run-time of each and start to replace ones whose
> run-time is less than what it should be.  With the low price of
> relatively good capacity, I have found it seldom useful to replace
> batteries.  The older UPSs are likely linear power supplies with large
> transformers etc.  New ones are far more efficient and give longer run
> times with the same amp hour capacity batteries.
> One person mentioned putting car or truck batteries in their UPSs to
> extend their run time.  That may not be a good solution.  The batteries
> in most UPSs are designed with slightly different chemistry and a
> different charging plan than the batteries used in automotive use.  I
> have found when trying this, that the automotive batteries die in a short
> time from the charging that UPSs do.  In most cases less than a year of
> life.  It has all to do with the charging.  I was told that the UPS
> battery uses a "float" charge.  I do not know what the difference is
> between that and what a standard charger for automotive batteries does.
> I look forward to hearing the experinces of others over this week-end.
> Regards
> Jim H
> James A. Haliburton
> On-Site Computer Services of Halifax
> Suite 100, 25 Walton Drive
> Halifax, Nova Scotia
> Canada     B3N 1X6
> Office/ Cell           : (902)499-5250
> Home/After Hours : (902)477-8342
> e-mail      : jim at on-site.ns.ca
> Please avoid sending me MS Office attachments.   For an explanation see
> this:
> http://www.nothingisreal.com/dfki/no-word
> _______________________________________________
> nSLUG mailing list
> nSLUG at nslug.ns.ca
> http://nslug.ns.ca/cgi-bin/mailman/listinfo/nslug

More information about the nSLUG mailing list