Re: 9406-720 Internal Battery Self Test

It would absolutely be valid to use it for that purpose.

Remember some things to consider however:

1) RUnning PWDDWNSYS *IMMED when the system has been sitting in restricted state will be EL-QUICKO compared to having 2000 jobs running, thousands of open files, and dirty RAID caches all around. You MUST understand how long it takes YOUR system to do PWRDWNSYS when running full song.
- Allow for this amount of time PLUS a good cushion.

2) Consider if you want to run your UPS down to it's last Watt or not. Think about what happens when the power comes back on, do you want some battery left so that it's OK to start the system back up fairly soon or not? Years ago my co-manager (who was in charge of hardware) would let our S/36 run until the UPS was crying for mercy and then shut it down. Then upon power restoration she literally hit the Power button the second the UPS came on line. One day we lost power again within 15 minutes and she didn't even make it into the Computer room before the UPS checked out. Ugh.

3) There are three signals available: UPS on Utility Power, UPS on Battery Power, UPS Batteries weak. When you go from Utility to Battery you might want to Hold Jobq's, outq's, and such to prevent new work from starting. Possibly notify people, all that. On Low battery you likely will be shutting things down but it's possible with your UPS that there's not enough time to shut down at that point. This is an important thing to know!! You might need to begin shutdown before that point.

4) Do you have a GenSet? ALL the rules are different then. Usually generators in under 15 seconds so if you don't get "Utility Power Restored' within that time frame of the 'On Battery' signal, it's likely not going to run. Most of my GenSet equipped customers have less than 5 minutes of UPS battery so in that case: "We've got a problem."

- Larry "DrFranken" Bolhuis

On 4/1/2011 3:59 PM, Jack Kingsley wrote:

After reading this, is anyone basing the findings on this to fine tuning a
UPS delay program(if one is in use), would it matter or not.

On Fri, Apr 1, 2011 at 3:26 PM, Vern Hamberg<vhamberg@xxxxxxxxxxx> wrote:

And remember, if you got the Powerware through your IBM BP, it's a
5-year warranty - saved our butts at another place I worked.

Vern

On 4/1/2011 12:29 PM, DrFranken wrote:

If you are talking about the batteries INSIDE the 9406-720 itself, those
will run for 7 seconds and then shutdown, even if the batteries are
spanking new. The remainder of the battery life is reserved to hold the
contents of the system's main memory up so that it can be copied to disk
during the next (abormal) IPL.

As I recall the Eaton Powerware unit doesn't have a specific battery
test. Even if it did, these tests are a 'best guess' based on a number
of seconds on battery.

If you want a full length battery test follow this procedure:
Bring the system down to restricted condition (ENDSYS) and let it
sit there for a few minutes. This will get the disk caches mostly
cleared out.
Note the value of QUPSDLYTIM and then change it to 99999.
Now pull the power to the Eaton and count the time it takes to run
out of battery. It's your call if you don't actually let it fail and
plug it back in at say 10% of battery but to know how long you really
have it's gotta die.
Once the Eaton unit powers off make sure the system does hang on
for 6 or 7 seconds to at least know the internal 720 batteries are not
totally dead.
Plug the UPS back in and let it charge for 30 minutes or more to
put some life back into the batteries.
IPL, and remember to change QYPSDLYTIM back to it's previous

setting.

- Larry "DrFranken" Bolhuis.

On 4/1/2011 10:57 AM, Martin Spencer wrote:

Hi All,

Is there a method to test the internal batteries on a 9406-720 without
taking it off the UPS and running it via the internal batteries, or, do

we

need to power down the UPS while the 720 is running and wait for it to

crash

out?

We need to do some failure testing on some systems that are connected to
UPS's but we are unsure of how healthy the UPS's are (they are Eton
Powerware 9325s, as specified by IBMs Econfig, and are 4 years old).

What we need to find out is should there be a power failure how long can
the 720 continue to run once the UPS battery has depleted. Obviously the

UPS

are set to send the signal to the AS/400 to begin a PWRDWN IMMED as soon

as

there is a power disruption.

Your thoughts, as always, are appreciated.

Tin.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: http://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxx
Before posting, please take a moment to review the archives
at http://archive.midrange.com/midrange-l.