27 May
2006
27 May
'06
5:03 p.m.
I have two of these:
http://sunsolve.sun.com/handbook_pub/Systems/SSA_219_7Slot/SSA_219_7Slot.ht…
Each has a pair of PSU's in it. I assume for redundancy.
However, the PSU's in one unit are strappable 110/220VAC
while those in the other unit are 220VAC. <frown>
If the PSU's are just redundant and aren't actively load
sharing, then I *should* be able to swap one 110/220VAC
for a 220VAC and run each array off a single supply
(taking care to hack the sense signal for the second
supply so the array doesn't continually report a supply
failure).
Am I missing anything here? Granted, the two supplies
possibly *do* share the load as a side-effect in normal
operation. But, I assume the intent was for a single PSU
to be capable of powering the entire device in the event of
a failure in the "other" PSU?
Thanks!
--don
28 May
28 May
10:06 p.m.
Don Y wrote:
I have two of these:
http://sunsolve.sun.com/handbook_pub/Systems/SSA_219_7Slot/SSA_219_7Slot.ht…
Each has a pair of PSU's in it. I assume for redundancy.
However, the PSU's in one unit are strappable 110/220VAC
while those in the other unit are 220VAC. <frown>
If the PSU's are just redundant and aren't actively load
sharing, then I *should* be able to swap one 110/220VAC
for a 220VAC and run each array off a single supply
(taking care to hack the sense signal for the second
supply so the array doesn't continually report a supply
failure).
Am I missing anything here? Granted, the two supplies
possibly *do* share the load as a side-effect in normal
operation. But, I assume the intent was for a single PSU
to be capable of powering the entire device in the event of
a failure in the "other" PSU?
Not quite the same array, but we've got a couple of StorEdge A3000 array
controllers on our pair of E6500 boxes and they certainly work as described -
later technology maybe, but it's an indication that your reasoning's correct.
They certainly can run off one supply; one of the dual supplies in one of our
units went into meltdown a while back and everything kept going happily on the
one supply (just the room filled with smoke). No idea why it failed, except
that it was first power-up after a few months being turned off, so possibly a
stress-related thing.
cheers
Jules
29 May
29 May
3:16 a.m.
Don Y wrote:
> Am I missing anything here? Granted, the two
supplies
> possibly *do* share the load as a side-effect in normal
> operation. But, I assume the intent was for a single PSU
> to be capable of powering the entire device in the event of
> a failure in the "other" PSU?
I'm a sysadmin who unfortunately works with these things every day...
Yes, the intent is failover. You plug both in, and if one fails, the
other carries the unit. IIRC, your unit will make an annoying beeping
sound if one of the power supplies goes south.
--
Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/
Help our electronic games project: http://www.mobygames.com/
Or check out some trippy MindCandy at http://www.mindcandydvd.com/
A child borne of the home computer wars: http://trixter.wordpress.com/
3:50 a.m.
Jim Leonard wrote:
Don Y wrote:
I'm sure I can workaround the alarm -- add a jumper between the sense
inputs (on the "environmental board") for the two power supplies.
But, I wonder just how much margin is designed into the supplies.
I.e. when you have a PSU failure, the faulty power supply is
usually *replaced*. So, the one supply that is carrying the
entire load eventually gets some relief.
Yet, I suspect most sites don't have cold spares for such
replacements -- so the PSU probably *usually* ends up running
the array for a week or more (?).
I'd just want to avoid accelerating a failure in the supply
due to the added load.
> Am I missing anything here? Granted, the two
supplies
> possibly *do* share the load as a side-effect in normal
> operation. But, I assume the intent was for a single PSU
> to be capable of powering the entire device in the event of
> a failure in the "other" PSU?
I'm a sysadmin who unfortunately works with these things every day...
Yes, the intent is failover. You plug both in, and if one fails, the
other carries the unit. IIRC, your unit will make an annoying beeping
sound if one of the power supplies goes south.
9:48 p.m.
Don Y wrote:
Actually, check on the back of the unit -- unless I'm thinking of the
wrong unit, there should be a pushbutton to cancel the alarm.
--
Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/
Help our electronic games project: http://www.mobygames.com/
Or check out some trippy MindCandy at http://www.mindcandydvd.com/
A child borne of the home computer wars: http://trixter.wordpress.com/
I'm a
sysadmin who unfortunately works with these things every day...
Yes, the intent is failover. You plug both in, and if one fails, the
other carries the unit. IIRC, your unit will make an annoying beeping
sound if one of the power supplies goes south.
I'm sure I can workaround the alarm -- add a jumper between the sense
inputs (on the "environmental board") for the two power supplies.
10:03 p.m.
Jim Leonard wrote:
Don Y wrote:
Actually, check on the back of the unit -- unless I'm thinking of the
wrong unit, there should be a pushbutton to cancel the alarm.
It's on the front. But, just as easy to add a jumper and
never have to worry about that! :>
I'm a
sysadmin who unfortunately works with these things every day...
Yes, the intent is failover. You plug both in, and if one fails, the
other carries the unit. IIRC, your unit will make an annoying
beeping sound if one of the power supplies goes south.
I'm sure I can workaround the alarm -- add a jumper between the sense
inputs (on the "environmental board") for the two power supplies.
30 May
30 May
12:14 a.m.
Don Y wrote:
Is that really a good idea? Failed power supplies should warrant
attention... The alarm is only supposed to sound when something FAILS,
not if something is MISSING.
--
Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/
Help our electronic games project: http://www.mobygames.com/
Or check out some trippy MindCandy at http://www.mindcandydvd.com/
A child borne of the home computer wars: http://trixter.wordpress.com/
Actually,
check on the back of the unit -- unless I'm thinking of the
wrong unit, there should be a pushbutton to cancel the alarm.
It's on the front. But, just as easy to add a jumper and
never have to worry about that! :>
12:52 a.m.
Jim Leonard wrote:
Don Y wrote:
Is that really a good idea? Failed power supplies should warrant
attention... The alarm is only supposed to sound when something FAILS,
not if something is MISSING.
Think about it... the whole point of this is to remove one
of the 110/220 power supplies from the array that has two
220V (only!) supplies and move it into that second array.
I.e. each array now has *one* power supply (since I am running
off of 110VAC). So, the second suply is already MISSING.
If we assume (the original question) that it is safe to run the
array on one supply, then this is OK for both arrays.
When / if this power supply dies, there is NO backup supply
for it. So, what does it matter if the alarm cant sound
when this (only!) supply fails? :>
Actually,
check on the back of the unit -- unless I'm thinking of the
wrong unit, there should be a pushbutton to cancel the alarm.
It's on the front. But, just as easy to add a jumper and
never have to worry about that! :>
1:17 a.m.
Don Y wrote:
Think about it... the whole point of this is to
remove one
of the 110/220 power supplies from the array that has two
220V (only!) supplies and move it into that second array.
Sorry, I missed that part :-)
--
Jim Leonard (trixter at oldskool.org) http://www.oldskool.org/
Help our electronic games project: http://www.mobygames.com/
Or check out some trippy MindCandy at http://www.mindcandydvd.com/
A child borne of the home computer wars: http://trixter.wordpress.com/
12:30 p.m.
Don Y wrote:
When / if this power supply dies, there is NO backup
supply
for it. So, what does it matter if the alarm cant sound
when this (only!) supply fails? :>
Depends on whether there are other alarm conditions, I suppose - critical
over-temperature or somesuch.
I don't think the later A3000's have an alarm - at least none went off when
one of the supplies in one of ours went bang and threw out smoke. Knowing Sun
it's software-configurable and someone had previously switched it off :-)
cheers
Jules
12:58 p.m.
On 5/29/2006 at 9:52 PM Don Y wrote:
When / if this power supply dies, there is NO backup
supply
for it. So, what does it matter if the alarm cant sound
when this (only!) supply fails? :>
Why is it necessary to disable the alarm? Why not simply make up a harness
to bridge the two power connectors with the same supply? If it goes, it
goes--otherwise, if it's working, it looks like both PSU's are operating.
...or am I missing something here?
Cheers,
Chuck
6:27 p.m.
Jules Richardson wrote:
Don Y wrote:
That's the advantage of just adding a jumper between the
sense input from PSU #1 (installed) and PSU #2 (removed)...
it tricks the alram circuitry into thinking both PSU's
are good without interfering with the reporting of any other
alarms.
If PSU #1 dies, then reporting an alarm is a moot point...
since the entire array will be dead (given that PSU #2
does not exist!)
When / if this power supply dies, there is NO
backup supply
for it. So, what does it matter if the alarm cant sound
when this (only!) supply fails? :>
Depends on whether there are other alarm conditions, I suppose -
critical over-temperature or somesuch. I don't think the later A3000's have an alarm
- at least none went off
when one of the supplies in one of ours went bang and threw out smoke.
Knowing Sun it's software-configurable and someone had previously
switched it off :-)
6:29 p.m.
Chuck Guzis wrote:
On 5/29/2006 at 9:52 PM Don Y wrote:
The alarm isn't "disabled". Rather, it is tricked into not
(*continually*!) signaling a failure in PSU #2 -- which is not
present. Note my comment:
When / if this power supply dies, there is NO
backup supply
for it. So, what does it matter if the alarm cant sound
when this (only!) supply fails? :>
Why is it necessary to disable the alarm? Why not simply make up a harness
to bridge the two power connectors with the same supply? If it goes, it
goes--otherwise, if it's working, it looks like both PSU's are operating.
...or am I missing something here? > I'm sure I can workaround the alarm -- add a
jumper between the sense
> inputs (on the "environmental board") for the two power supplies.
6633
days inactive
6636
days old
12 comments
4 participants
participants (4)
-
cclist＠sydex.com -
dgy＠DakotaCom.Net -
julesrichardsonuk＠yahoo.co.uk -
trixter＠oldskool.org