HP ML350 G5 internal & external health leds blinking on front panel

[Glenn _G]

Hi.
I've got a ML350 G5 which shuts down at random without any useful information in the iLO logs or the system management homepage.
Both the internal and external health LED on the front panel is blinking red when this happens.
If I re-insert the power cord the server works again, but after some hours or days it shuts down again.

What does the two blinking red lights mean?

[juan quesada]

check pages 73 and 78
http://bizsupport1.austin.hp.com/bc/docs/support/SupportManual/c00708990/c00708990.pdf

if you have both red, start by Troubleshooting the external led (power subsystem) since the internal depends upon it

BTW, very important, this server consumes a lot of Watts, where is it connected to?

[Glenn _G]

When the server powers off some LEDS on the mainboard is amber, so page 78 is useful. Thanks.

The server is connected to a 3000VA UPS.
One other identical server is connected to the same UPS.

BTW, the power supply in the failing server is only 4 months old.

[Taifur]

Hi Glenn,

As, external health led blink red that means you have to check power cable, power supply.
and
As internal health led blink red that means, you have to open the server cover and check all of spare i.e processor, RAM, systemboard.
If any led blink red besides of any spare ie.RAM, processor, system board then you have to understand that this is faulty.
Then replace faulty spare and check the server status.

Rgds//
Taifur

[juan quesada]

what is your power supply revision?
should be something like 02n, 03n with a green or yellow or red dot close to the revison number, 04n 06n can also be there

[Glenn _G]

The power supply is REV 03M. It was a blue dot next to the revision number.

I tried to connect the power supply to one of the other outlets on the UPS, still got the same problem.

I will now try to connect it directly to the wall outlet with a new cable.

LED numer 8 is yellow when this error happens.
From what I can see on page 78 in the manual (link posted earlier) this is PPM 2 failed.
But there is only one CPU in this server.
It COULD be LED 7 that is yellow, but that means CPU FAN 2 failure... Which of course is not installed.


On page 77 of the manual there is a system maintenance switch.
If I put this in the ON position, will the RAID configuration be erased?

[Glenn _G]

Im referring to Switch number 6. Clear NVRAM.

[juan quesada]

if your revision is 03m with blue dot and you have an amber led next to PPM 2 without nothing on the socket, call HP and have them replace your power supply, 90+ % of chances it will fix your issues...been there,done that.

[juan quesada]

BTW, clearing NVRAM from the dip switch in the system board will not delete your RIS copies.

[HP ML350 G5 internal e]

Glenn,

I also have the same issue with ML 350.How you resolved this?

[Glenn _G]

I have not resolved this problem.

[Georges Bolssens]

Same problem started occurring here a couple of weeks ago with a 19" Rackmount ML350 G5. I'm expecting a new power supply today. I'll keep you posted on the outcome.

Best regards,
GB

[berni_1]

Hi!

I have the same problem, did you already solve it?

[Spawn_BR]

I'm having the same problem with this server. ML 350
With the tests I hope that is the source, I am providing one.
Someone solved?
Grateful

[Lincy_Zachariah]

Are you having issue with ML350 G5 server.

Were there any changes done to the server recently?

Any other performance issue with the server like unexpected restart or shutdown?

Please refere the the page 73, 74 and 78 for the LED indications.

Thanks,

[mastsrl]

I experienced the same problem on a server when doing testing since it restarts alone.

After running a complet einsight diagnostics overnight, the following day i had this exact same issue.

All firmwares are the latest, IML and lo2 log don't show ANYTHING about any errors, other than server power removed

PSU is rev: 06M yellow dot

and it's waaaaaaay out of warranty so i doubt HP will do anything

is there some advisory or anything regarding this psu problem?, this server has been running for 45months nonstop with no issues before...

[Aunap]

Hi, 

I have the exact same problem.

Has anyone solved this?

Is HP doing anything?

Cheers

[NelsonNVTech]

I had the same issue. The array confoguration utlilty had an alert about the battery for the E200i controller card needing to be replaced. I have replaced the battery and am monitoring the server. 

Could this cause the 2 red lights and then no response from the server? 

P.S. The hard drive activity lights are also dead when the 2 red lights turn on.

[mastsrl]

Aunap,

"no and no"

problem was never solved, we tried swapping PSUs with other server (different model, same generation) problem persists and makes the server useless.

Server was essentially discarded, HP gave no support whatsoiever(being way out of warranty), there are no advisories or anything and everything essentially points to a fauilty motherboard which neither we or the customer want to change (only to discover the problem was the PSU converter board for example) as it costs more than buying a new server.

We also tried removing ALL boards, completely disconnecting the drive cages too(power+signals), removing all ram modules, putting in different(brand new known-good HP sticks), changing CPU, nothing, server FAILS in the same manner at complete random times regardless of load or task done

[Paulie_M]

NelsonNVTech,

did the replacement of the e200i battery solved the issue?

[BirdmanEccles]

I had a problem with an HP ML350 Gen 5 tower.

The symptoms were:- Both front panel LEDs flashing red when power button pressed - NOTHING else.

Internally the PPM2 LED was an amber colour. 

Disconnecting and reconnecting the mains made the front panel and board LEDs look normal but when the power button was pushed the above symptoms happened again!

I then, based on the fact that the external health LED on the front panel was flashing red and this issue would need solving before moving on, took a power supply from a redundant server (in my garage for the last 6 months!) and replaced the power supply in the problem server. Checked everything was tied down, connected properly etc., and reconnected the mains. Guess what - the server started up and went through the motions fine.

I would therefore say that the first thing to try if you have the symptoms above is to replace the power supply - after checking that the wattages, and preferably HP part nos are the same (I was lucky the power supply I used was exactly the same in these respects.

[edonley]

We had the same issue with an ML350 G5 tower model.  Would shut down randomly.  Tried several different fixes with no results.  Luckily we didn't need a server class machine to fill that role anymore.  Once I had it replaced and down to my office I cracked it open and noticed all the capacitors on the top of the system board were swollen, several had actually broke and were leaking.  I counted ten of them that were like that.

[Fred_39]

We have four ML350 G5 servers. Two of which are working and two of which are dead with exactly the same problem: both internal and external LED's blinking red and the amber PPM2 LED flashing.

We replaced the PSU and backplane on one of the servers and it made no difference at all. It was still dead with the same problem. When the second server died today with the same issue I Googled and found the following post:

http://h20565.www2.hp.com/portal/site/hpsc/template.PAGE/public/kb/docDisplay/?spf_p.tpst=kbDocDisplay&spf_p.prp_kbDocDisplay=wsrp-navigationalState%3DdocId%253Demr_na-c03562080-1%257CdocLocale%253D%257CcalledBy%253D&javax.portlet.begCacheTok=com.vignette.cachetoken&javax.portlet.endCacheTok=com.vignette.cachetoken 

and this lead me to replace the cpu0 processor with one out of the second dead machine we have and bingo! it leapt in to life.

However, it isn't the end of the story as I then took the second processor out of one of the working servers and put it in the cpu0 slot of our original dead machine and it still won't start and still has the same flashing LED's.

What is going on with these servers!

[Paul Daws]

Your servers may not both have the same fault.

So where a CPU may fix one machine it may not fix the other machine.

I'df suggest you try replacing the PPM next, this may fix the problem.

[DIMITRIS_10]

hI Paul

I have the exact same problem

Tried all of the above, haven't replaced the motherboard yet (and I m not going to)

What do you mean by PPM?