My 4500 doesn't respond to the power control switch

[LarryInCa]

Same server, different problem

I can't get my 4500 server to wake back up after powering it down. The last action I took was to go into the GUI and shut it down, as I needed to move the box.

After plugging it back in, nothing happened while pressing the power button

Tried swapping power cords, but no change. The green led on the PS blinks when I plug it back in, and the MB has a green LED in the center that lights. Remove the cord and both of the LED's go out.

I checked all of the internal plugs to make sure they were seated, and they were. I unplugged and replugged the ribbon cable that goes to the power switch / led PCB. I pulled the CMOS battery, and replaced it. I pulled the PS and reseated it.

No change.

Could use some suggestions on what to try next.

[bitor]

Hello LarryInCa,

Not for sure what it could be exactly. It could be your power supply is weak/strained/going bad which I think may be the case. Do you hear the drives spin up as you power up the server? Anyway, something to consider as you trouble shoot.

bitor

Originally Posted by LarryInCa

Same server, different problem

I can't get my 4500 server to wake back up after powering it down. The last action I took was to go into the GUI and shut it down, as I needed to move the box.

After plugging it back in, nothing happened while pressing the power button

Tried swapping power cords, but no change. The green led on the PS blinks when I plug it back in, and the MB has a green LED in the center that lights. Remove the cord and both of the LED's go out.

I checked all of the internal plugs to make sure they were seated, and they were. I unplugged and replugged the ribbon cable that goes to the power switch / led PCB. I pulled the CMOS battery, and replaced it. I pulled the PS and reseated it.

No change.

Could use some suggestions on what to try next.

[LarryInCa]

Unfortunately, no obvious signs of life other than the LED's.

I was poking around in the overland support area, and came across a FAQ about the power button not working to power the machine down, but nothing about powering it up.

http://support.overlandstorage.com/j...earchPage=true

Could it be something as simple as a bad power switch? If so, can I just short out a couple of pins somewhere to simulate the button being pressed?

[bitor]

Hello again LarryInCa;

The link doesn't fit your problem. MAYBE it could be as simple as a power button. One way of testing the power supply is to remove 2-4 drives(keep track of where they were) and try to power up the unit with one drive. If you have it in a RAID it will bring up the recovery console via the webgui. Starting up a RAID without 3 drive not being in the server MAY cause a problem with your data on the drives and the RAID.
So do this at your own risk.

I hope you have your data backed up.

bitor

Originally Posted by LarryInCa

Unfortunately, no obvious signs of life other than the LED's.

I was poking around in the overland support area, and came across a FAQ about the power button not working to power the machine down, but nothing about powering it up.

http://support.overlandstorage.com/j...earchPage=true

Could it be something as simple as a bad power switch? If so, can I just short out a couple of pins somewhere to simulate the button being pressed?

[Phoenix32]

Originally Posted by LarryInCa

I was poking around in the overland support area, and came across a FAQ about the power button not working to power the machine down, but nothing about powering it up.

Larry, forget this power supply crap (remember what I said in the previous post). While it is "possible", it is very unlikely and not what the indications here indicate.

You are on the right track with the power button. The 4400/4200/4500/15000 etc are notorious for the power button going bad. What most often happens is the button gets pressed too hard and the little switch seperates from the circuit board it is soldered to. Check this first!

[LarryInCa]

The 4400/4200/4500/15000 etc are notorious for the power button going bad. What most often happens is the button gets pressed too hard and the little switch seperates from the circuit board it is soldered to.

The switch looks okay from a visual inspection. I applied light pressure, and the switch doesn't appear to move.

And thinking back to the problem that first brought me to this forum, I recall trying to use the power button to shut the unit down, and it didn't work. Since I couldn't get into the unit anyways, I just pulled the plug to shut it down. And it came back up once the power was restored (back up meaning it would boot into something).

Which brings be back to the power switch. The only way that I know of to confirm if it's the switch or something else is to swap in a known-good switch/led assembly. Any pointers on where I can find one?


Thanks,

Larry

[blue68f100]

If you dumped the bios settings did you restore the settings back to factory spec??????

May try using a jumper across the switch and see it starts.

[willPower]

.

[Max8]

Originally Posted by Phoenix32

You are on the right track with the power button. The 4400/4200/4500/15000 etc are notorious for the power button going bad. What most often happens is the button gets pressed too hard and the little switch seperates from the circuit board it is soldered to. Check this first!

I had one of these recently the 4500 unit was very inconstant but as soon as I changed over the switch block it has worked without issue...

The switch came from a faulty 4200 with a dodgy m-board or backplane that I have…

I guess they cousl be a little more robust but it comes down to the user's temperament… Say no to reckless IT Hardware abuse!

[LarryInCa]

Originally Posted by willPower

Have you tried removing the front bezel and pressing the little black button (on the chassis) that sits directly behind the power button (on the bezel)?

Tried that multiple times, both before and after reseating cables, power supplies, memory, drives, MB connector cables, CMOS battery, etc ...

The frustrating thing is that this happened only after an orderly shutdown initiated in the GUI. If I had just pulled the plug, I'm sure the box would've come back up once power was returned

[LarryInCa]

Originally Posted by Max8

The switch came from a faulty 4200 …

So this switch is used in multiple units? That's good to know, as it increases the chances of finding a replacement part.

Anybody know where I might find one? I sent an email to Overland Support yesterday asking if they sell parts, and I'm waiting to hear back from them. I might even pick up the phone and give them a call if I don't get a reply.

[LarryInCa]

Back up and running after replacing the PCB that holds the switch.

[Max8]

Originally Posted by LarryInCa

Back up and running after replacing the PCB that holds the switch.

Hi Larry,

I am sure we are all glad to hear that you are back up and running!

Did you get the replacement from Overland?
If not where did it come from?
How much did it set you back?

Thank you in advance for the info...

Matt

[LarryInCa]

Sorry Matt, didn't mean to be cryptic.

I was able to get the part from Overland for less than $20. It took some persistence through a Support ticket with my expired maintenance to get to somebody that could help. They don't appear to have an "order online" capability (at least none that I was able to find).

The part is labeled "FRU, Power/Reset Switch PCBA" with a part number of 5325302181. The installation instructions say this will work in a 4200 / 4500 / 15000.

Easy access to change the part, and all that is needed is a good, small, Phillips screwdriver to remove the old part, and a small flat blade screwdriver to help encourage the ribbon cable to separate from the circuit board(s).

[achapman]

A further note on this. I've just bought a 4500 which has the same fault. Initially it would power up as the power was connected and pressing the power switch wouldn't power it down. Lately it won't power on and again the switch doesn't work.

I removed the switch assembly from a Snap 4400. This is very similar, but not identical. Effectively it appears that the 4500 switch board is an extended 4400 part with the addition of a white reset switch to the left of the black power micro switch. This means that the board is longer and the left mounting screw sits about 10mm further to the left.

As well as the power button working correctly, the two network LED's work as does the system LED. The power on LED doesn't illuminate though, so as a temporary fix it is usable.

The cable and connector that connects it to the disk backplane is identical to the 4500 switch board. I therefore tried the 4400 switch assembly on the 4500 and it works. The 4500 will now power up and down, confirming the 4500 switch board is faulty.

Since I'd rather have my 4500 working than my 4400 I may leave the switch board on the 4500 for now. I imagine getting a spare part may be tricky as I haven't a service contract with Overland.

[Max8]

These switches are delicate and when dudes who are impatient loose it they can take it out on the power on button, I have recently seen them work intermittently as well... its just a painfull thing; everyone is too used to hard resetting windows...

Having the software maintenance may help you, it will at least give you access to GOS but if you start with the 1st level support people you probably won’t get that far…

The easiest course of action will be to locate a failed or incomplete Snap that is compatible...

If you cannot locate one or you prefer a new part your biggest issues will be;
a) getting a local overland distributor or a contact at overland to assist you..
b) stocks of the part may have been exhausted with be the age of these it is quite possible

Contact the local distributor and see if you can get someone on the phone who is into SnapServers and wants to help i.e. call and ask for the SnapServer "guru", it may be a tech or a product manger
If they cannot help you they may be able to talk to overland for you or at least introduce you…

LarryInCa tells us that; The part is labelled "FRU, Power/Reset Switch PCBA" with a part number of 5325302181. The installation instructions say this will work in a 4200 / 4500 / 15000.

[bitor]

LarryInCa, Thanks for the information.

[Phoenix32]

Originally Posted by bitor

LarryInCa, Thanks for the information.

Yeah, because you were leading him on a wild goose chase when there was a simple and obvious solution, as usual...

[Phoenix32]

Originally Posted by achapman

The 4500 will now power up and down, confirming the 4500 switch board is faulty.

Can you solder? I have found that 70% to 80% of the time, the switch isn't bad, but the solder joints have come loose (even if they dont look like it). Sometimes the switch is bad, but worth a shot to try and resolder the switch joints to see for sure. Just be careful,it is delicate soldering.

Also, if it is for sure the switch button itself, you could remove the reset switch and put it in place of the power the switch for a simple temp fix until you can get a new board.

[bitor]

There were no simple obvious solutions to this problem. You are nuts. It's called trouble shooting. It's only an "obvious solutions" to those who know or experienced the problem. And to make it clear, before this thread, there were no solutions posted on the switches going bad on this forum concerning the units in question. And very little postings of trouble concerning theses switches in this forum.

Originally Posted by Phoenix32

Yeah, because you were leading him on a wild goose chase when there was a simple and obvious solution, as usual...

[Phoenix32]

Originally Posted by bitor

There were no simple obvious solutions to this problem. You are nuts. It's called trouble shooting. It's only an "obvious solutions" to those who know or experienced the problem.

Oh my goodness, THANK YOU BITOR!

No, I am not kidding. As much as I don't like you, I have to thank you for that one. That was as funny as reading the humor sections of the readers digest. I loved it!


Okay, okay, I am sure you don't get it, so I will explain it to you. But only because you tickled my funny bone...


1. You are right, there are no simple or obvious solutions to a technical problem for those who are not properly trained (which is what I have been saying about you for a long time now). i.e. Most automotive problems these days are not obvious to a person not trained as a mechanic. There are always exceptions, but you have proven not to be one of them over and over.

2. Nuts? Well, that remains to be seen, but is always possible I suppose. I do see a VA (Veterans Affairs) psychiatrist regularly for my PTSD (Post-traumatic stress disorder), but in their highly trained opinions, they don't think I am nuts. But you obviously know more than they do, right?

3. Troubleshooting? Well, yes. Troubleshooting is how a problem is found, but there is a proper order to doing that. You have a tendancy to just jump in the middle and search in a random order, rather than a methodical and logical process from beginningto end.


Okay, okay, so you still don't get it. Let me try this simple example to see if you will understand that.

A lady is driving down the road. Her car starts sputtering, then after a little while dies on the side of the road.

a) She calls a friend who thinks he kows about cars (like you and the subjects on this forum). He (you) says, "You need to have your car towed, then have the fuel pump tested, along with the injection system, and a full set of diagnostics on the engine via its computer chip". i.e. Kind of like checking the power supply and pulling drives etc for the problem described in this message thread.

b) She calls a real mechanic who knows proper troubleshooting and common problems related to general auto mechanics. He says, "Ma'am, is it out of gas?" which it is.

While it could have been the fuel pump, injection system or some other deep rooted problem in the engine, it was proper troubleshooting procedure to check the obvious and/or simple things first. i.e. Mechanical switches are a simple and very common failure point in electronics and in this case, would be a standard starting point for troubleshoting by a trained technician.


Get it yet Bitor? Dumb ass... Learn what you yourself are doing before you try to teach others. Listen to those who are far better trained and educated on the subject than you and maybe you will improve yourself and learn something.

And pleeeeeaaase do not try to compare your pitiful knowledge level about electronic hardware to many of those like myself here on the forum. I have many years (near doctorate level) of formal education, training, and in the field experience in electronics engineering. And many others here have similar levels of formal knowledge. Call me or others megalomaniacs or whatever else you want, but it doesn't change you are out gunned and out classed in this field here.

BUT!

I was dead serious before. It was funny as hell to read your response and I enjoyed it very much. Thank you. In afterthought, maybe I should just let you make a fool of yourself and just enjoy it.

[Max8]

Hello Gents...

I will definitly try to re-solder the faulty one I have, see if I can get it happening again...
I don't like soldering but I have managed to repair the same tv remote, 3 times now, it keeps coming loose but at least it works for a while...

[achapman]

I ended up getting a switch assembly off a Dell 725N. Strangely the 725N board, although exactly the same size, same connectors etc doesn't have the reset switch, only the on/off switch. Apart from that it just plugs in and works fine, so I now have a fully working 4500 again.

Thanks for the soldering suggestion - but my soldering skills are more suited to electrics (household) than electronics. I don't think the board would survive them.