[SOLVED] Asus Z10PE-D8 WS - unstable server tortures me since a few days

[gosha1100]

Hey guys,
I’m a developer and kinda new to the whole hardware stuff.

I build a Server with a Asus Z10PE-D8 WS and two Xeon E5-2643 v3. A gtx 980 Ti is in the 2nd PCIe slot.
1000W be quiet PSU

The whole thing is kinda weird.
When I turn it on, it runs for 2 secs, turns off, turns on again and then I get a asus screen with a beep sound. It wants me to enter the setup to recover the BIOS setting.
Sometimes the BIOS settings are saved sometimes not.
I think I enabled secure booting or something and
at some point I managed to let it boot my Ubuntu boot stick and I did the whole Ubuntu installation but at the moment where u have to get the boot stick out and press enter to reset, I entered a passive black screen state.
I turned it off and the BIOS settings are at the state of the beginning.
The biggest issue is this passive black screen state.
Sometimes I get the asus screen with the beep sound, sometimes just the black screen.
This whole thing is so inconsistent and weird.

I already updated my BIOS to the current version.
I’ve read that the CMOS battery could be the problem, but it isn’t that old tbh.
I’ve also read something about RAM problems.
I put 8 x 8GB RDIMM inside.



Thanks for your help

 

[Ralston18]

What OS is installed?

Update your post to include full system hardware specs and OS information.

PSU: make, model, wattage, age?

Disk drive(s): make, model, capacity, how full?

Start with a new replacement CMOS battery as a matter of elimination.

"Not that old" being very relative. Could be a cheap/substandard or faulty battery.

When you built the server what thermal paste was used and how was the thermal paste applied?

 

[Darkbreeze]

Did all your memory come together in one kit or is it all made up of DIMMs from multiple kits even if they are the same model? If it is different models, please list each different model and number of DIMMs?

The C612 chipset released in 2014 so your CMOS battery is WELL past any date where it could potentially be suspect. I'd either replace or test it because it is the most probable cause without further information.

It might also be a good idea to test each CPU individually to see if perhaps one of them is to blame and check each motherboard carefully to ensure there is not even one bent pin.

This system is like 9 years old so don't be terribly surprised if it ends up simply being a motherboard issue either. I'd imagine if this is running the original power supply, that is likely suspect as well. What is the model of the current one?

 

[gosha1100]

What OS is installed?

Update your post to include full system hardware specs and OS information.

PSU: make, model, wattage, age?

Disk drive(s): make, model, capacity, how full?

Start with a new replacement CMOS battery as a matter of elimination.

"Not that old" being very relative. Could be a cheap/substandard or faulty battery.

When you built the server what thermal paste was used and how was the thermal paste applied?

There is no OS installed and the components are probably from 2018. I’m planning to install Ubuntu on it.
I just put the cooler on it without paste, because I wanted to test some RAM sticks.
I checked in the BIOS they didn’t overheat at all, so I thought it‘s ok.
initial there were 2 western digital NAS HDDs. I swapped one with a Samsung 860 PRO SSD, because I thought it will maybe fix it.

 

[gosha1100]

Did all your memory come together in one kit or is it all made up of DIMMs from multiple kits even if they are the same model? If it is different models, please list each different model and number of DIMMs?

The C612 chipset released in 2014 so your CMOS battery is WELL past any date where it could potentially be suspect. I'd either replace or test it because it is the most probable cause without further information.

It might also be a good idea to test each CPU individually to see if perhaps one of them is to blame and check each motherboard carefully to ensure there is not even one bent pin.

This system is like 9 years old so don't be terribly surprised if it ends up simply being a motherboard issue either. I'd imagine if this is running the original power supply, that is likely suspect as well. What is the model of the current one?

Let me check the psu is kinda hidden in the case. And I also check the DIMMs. And thanks for the CMOS hint guys. I thought that it‘s not that old xd, I would’ve probably swapped it in 2 weeks after thinking about everything else.

 

[gosha1100]

Let me check the psu is kinda hidden in the case. And I also check the DIMMs. And thanks for the CMOS hint guys. I thought that it‘s not that old xd, I would’ve probably swapped it in 2 weeks after thinking about everything else.

The psu is a be quiet straight power 11 1000W
The DIMMs are all the same: CT8G4RFS4213.18FA2

 

[Darkbreeze]

Did you buy that PSU and install it or did it come with the system?

Has this system ever run since you've had it or is this the initial attempt to get it working after obtaining the machine and parts?

 

[gosha1100]

Did you buy that PSU and install it or did it come with the system?

Has this system ever run since you've had it or is this the initial attempt to get it working after obtaining the machine and parts?

Tbh I don’t think that most of the components were used, they look brand new, but they are for sure pretty old. Probably around 2016 or something. I think all the components are pretty much the same age and were meant to be a build.
The company I‘m currently working at, bought an agency and I was allowed to keep things I can need. I found all those things in the Server room and build it at home.
So the hint with the CMOS battery is most likely the issue.
I just figured out that CMOS batteries doesn’t last that long if you don’t use the system. I thought it would be the the other way around. 💀

 

[Darkbreeze]

Well, that's pretty decent memory and it should be compatible with that board and CPU. So, that is unlikely to be the problem. It might not be a bad idea though to try only a single DIMM, or like four DIMMs, and make sure you are populating them according to the motherboard population guidelines because many quad and higher channel configurations have some weird population rules sometimes. They are definitely not all exactly the same like consumer dual channel boards are.

Definitely try the CMOS battery. That's probably the problem but if not let me know and we'll look at some other things. PSU is fairly decent, but if it's as old as everything else it's likely about right at it's five year warranty expiration and once you get things right you might want to seriously consider replacing it with a new high quality unit.

At about five years, most CMOS batteries will either be beginning to become weak or even might be dead especially if they were sitting on the shelf for a while before being installed in their motherboard and shipped out. I've seen motherboards not even a year old with weak or dead CMOS batteries. I can only assume that some stock that had been stored for a while was used during assembly at the factory. Been a while since I've seen that happen but I have seen it happen so I never assume the CMOS battery is good if there are symptoms. I always replace or at least test with a multimeter.

 

[gosha1100]

Well, that's pretty decent memory and it should be compatible with that board and CPU. So, that is unlikely to be the problem. It might not be a bad idea though to try only a single DIMM, or like four DIMMs, and make sure you are populating them according to the motherboard population guidelines because many quad and higher channel configurations have some weird population rules sometimes. They are definitely not all exactly the same like consumer dual channel boards are.

Definitely try the CMOS battery. That's probably the problem but if not let me know and we'll look at some other things. PSU is fairly decent, but if it's as old as everything else it's likely about right at it's five year warranty expiration and once you get things right you might want to seriously consider replacing it with a new high quality unit.

At about five years, most CMOS batteries will either be beginning to become weak or even might be dead especially if they were sitting on the shelf for a while before being installed in their motherboard and shipped out. I've seen motherboards not even a year old with weak or dead CMOS batteries. I can only assume that some stock that had been stored for a while was used during assembly at the factory. Been a while since I've seen that happen but I have seen it happen so I never assume the CMOS battery is good if there are symptoms. I always replace or at least test with a multimeter.

Aight I will try the CMOS first. There’s CR2032 written on it, so I think any CMOS with this number should be fine I guess.
Thanks for the help. You guys told me things I would’ve never thought about. 👾
I‘ll keep You updated, hopefully the CMOS will solve the problem.

 

[Darkbreeze]

Yes, pretty much all desktop motherboards use the CR2032 CMOS battery.

 

[gosha1100]

Yes, pretty much all desktop motherboards use the CR2032 CMOS battery.

So I swapped the CMOS and the BIOS finally keeps all changes saved, which allowed me to boot up into the ubuntu environment. The only problem now is booting up consistently. Sometimes I get a black screen, sometimes I get into ubuntu with no problem at all.
So you told me that I should try a single DIMM or 4 DIMMs right? I'll try that first, but I'm kinda unsure which slots are the right ones.
Luckily I'm planning to build a new gaming pc, so I can try it with a new PSU in some weeks.

 

[Darkbreeze]

FIrst thing to determine is what the model of the motherboard is?

It should be printed somewhere directly on the motherboard itself.

I know also that you said you updated the BIOS to the current version, but sometimes that isn't always accurate. So stating what the currently installed BIOS version is would be far more valuable. Often. like one in five times, when somebody says "it's on the latest version", it's not. And more than a handful of times I've seen people with semi-functional systems who though they updated to the right version, and turns out they used a version meant for a different revision of that motherboard, like, for example only, Z170-A v1.0 BIOS of some flavor on a Z170-A v2.0 board that uses entirely different firmware because the design of the board is not the same. So, always good to be TOTALLY sure on that aspect by stating the current BIOS version that is installed.

And, sometimes, they are simply mistaken and it is not the latest version even though it IS the correct board model and revision.

And I really would remove both of the CPUs, check the board to make sure there is not even one single bent pin on either of the pinbeds and that both CPUs are correctly oriented in the socket with the alignment marks on the CPUs matching the alignment marks on the socket/motherboard. Then check the manual to see which of the two CPU sockets is intended to be used if only one CPU is used, and try just one CPU, in that socket. If you still have the same results, try the other CPU in that socket. You will probably want to get a bottle of 91% isopropyl alcohol and some thermal paste as you might be taking the CPU in and out of the sockets a few times before all is said and done, unless you are already set up in that regard. Don't use 70% iso alcohol, it is, well, it CAN work, better than nothing, but really it is not strong enough to cut thermal paste for a good cleanup. Be sure also to use lint free cloths or paper coffee filters for cleaning. You do not want to leave bits and pieces of cloth, however microscopic they might be, on the surface between the CPU heat spreader and the bottom of the CPU cooler heatsink.

Using only one CPU and one DIMM will also have the additional benefit of MAYBE saying something about the capacity, given it's age, of the power supply. A lot of times a weak PSU can't support a lot of devices including two to four sticks of memory, extra drives, multiple CPUs, higher power draw PCIe devices like graphics cards or even regular additive devices like NIC cards or PCIe M.2 adapters, so if you see an immediate change with a single DIMM and a single CPU, then it becomes more probable that it's maybe PSU related.

 

[gosha1100]

FIrst thing to determine is what the model of the motherboard is?

It should be printed somewhere directly on the motherboard itself.

I know also that you said you updated the BIOS to the current version, but sometimes that isn't always accurate. So stating what the currently installed BIOS version is would be far more valuable. Often. like one in five times, when somebody says "it's on the latest version", it's not. And more than a handful of times I've seen people with semi-functional systems who though they updated to the right version, and turns out they used a version meant for a different revision of that motherboard, like, for example only, Z170-A v1.0 BIOS of some flavor on a Z170-A v2.0 board that uses entirely different firmware because the design of the board is not the same. So, always good to be TOTALLY sure on that aspect by stating the current BIOS version that is installed.

And, sometimes, they are simply mistaken and it is not the latest version even though it IS the correct board model and revision.

And I really would remove both of the CPUs, check the board to make sure there is not even one single bent pin on either of the pinbeds and that both CPUs are correctly oriented in the socket with the alignment marks on the CPUs matching the alignment marks on the socket/motherboard. Then check the manual to see which of the two CPU sockets is intended to be used if only one CPU is used, and try just one CPU, in that socket. If you still have the same results, try the other CPU in that socket. You will probably want to get a bottle of 91% isopropyl alcohol and some thermal paste as you might be taking the CPU in and out of the sockets a few times before all is said and done, unless you are already set up in that regard. Don't use 70% iso alcohol, it is, well, it CAN work, better than nothing, but really it is not strong enough to cut thermal paste for a good cleanup. Be sure also to use lint free cloths or paper coffee filters for cleaning. You do not want to leave bits and pieces of cloth, however microscopic they might be, on the surface between the CPU heat spreader and the bottom of the CPU cooler heatsink.

Using only one CPU and one DIMM will also have the additional benefit of MAYBE saying something about the capacity, given it's age, of the power supply. A lot of times a weak PSU can't support a lot of devices including two to four sticks of memory, extra drives, multiple CPUs, higher power draw PCIe devices like graphics cards or even regular additive devices like NIC cards or PCIe M.2 adapters, so if you see an immediate change with a single DIMM and a single CPU, then it becomes more probable that it's maybe PSU related.

Ok thank You, I will try all these things.
There is one thing I noticed.
When I cold start the system, it will boot up almost every time.
Just after running it for a while maybe like 30 mins. When I turn it off and turn it on, it sometimes has this booting problem.
I‘ll try and check all these thing and give you an update, thanks for the help

 

[Darkbreeze]

Ok, so lets be clear. When you are turning it off and back on, you are using the start menu shut down or restart options right? I know that seems like "well, duh" but you'd be surprised at some of the very strange things we see here all the time.

So, as long as you are not powering it off by holding down the power button or flipping off the switch on the back of the PSU or any other method that does not involved clicking "Shut down" or "Restart" on the start menu, then we don't have to backtrack.

One thing I would really like you to clarify is that you say "after running it for a while maybe like 30 mins" but previously you stated there is no OS installed, so if there is not an OS installed (Windows, Linux, or some other operating system) then what is it that you, or it, is doing for that 30 minutes? Are you just booting into the BIOS and that's it? I think some additional clarity on exactly what the problem is that you are actually having is and when it exactly it does and does not happen because right now I'm back to not really being sure.

 

[gosha1100]

Ok, so lets be clear. When you are turning it off and back on, you are using the start menu shut down or restart options right? I know that seems like "well, duh" but you'd be surprised at some of the very strange things we see here all the time.

So, as long as you are not powering it off by holding down the power button or flipping off the switch on the back of the PSU or any other method that does not involved clicking "Shut down" or "Restart" on the start menu, then we don't have to backtrack.

One thing I would really like you to clarify is that you say "after running it for a while maybe like 30 mins" but previously you stated there is no OS installed, so if there is not an OS installed (Windows, Linux, or some other operating system) then what is it that you, or it, is doing for that 30 minutes? Are you just booting into the BIOS and that's it? I think some additional clarity on exactly what the problem is that you are actually having is and when it exactly it does and does not happen because right now I'm back to not really being sure.

After swapping out the CMOS I was able to install desktop ubuntu and I just let it run for a while.
I tested it. When I restart in the OS it will boot up. So I flipped the PSU switch and now the system won't boot up.
I think it will boot up, when I let it be turned off for a while.

edit: yes works fine again.
edit: shutting it down from the start menu causes the same problem as well, restarting was no problem.

 

[Darkbreeze]

Is this board installed in a case or rack, or is it benched on a box, cardboard or something similar? What I'm wanting to know is if there is a power switch connected to this like there would be in a server rack or PC case, or if you are triggering it on using a screwdriver?

 

[gosha1100]

Is this board installed in a case or rack, or is it benched on a box, cardboard or something similar? What I'm wanting to know is if there is a power switch connected to this like there would be in a server rack or PC case, or if you are triggering it on using a screwdriver?

It's in a case. I just used the power switch from the case and it also started sometimes, when I flipped the PSU switch on.
All fans spin(case fans, GPU) and the case lights turn on. when I get a black screen keyboard and mouse lights won't turn on.

 

[Darkbreeze]

PSU switch should never "turn the computer on". If it does, there's a problem. In fact, the switch on the PSU should NEVER be turned OFF unless you are going to be working inside the case or replacing hardware, or there is an electrical storm in the area (In which case it should ALSO be unplugged completely from the wall as switching it off will stop nothing if there is a strike close enough to be damaging) or when doing a hard reset procedure. Otherwise, there is zero reason for having the PSU switch off under any normal conditions. And, if you ever flip the switch on the PSU on and the system immediately tries to power on without using a power switch or jumping the pins on the motherboard, then there is a REAL problem with something because it should never do that.

So, I'm almost tempted to say there is an issue with the power switch or something else on the case. I'd recommend you disconnect ALL of the wiring and connections coming from the front of the case to the motherboard and try powering on the board using the pwr pins as explained here.

View: https://www.youtube.com/watch?v=FuPZlliGqBw

 

[gosha1100]

PSU switch should never "turn the computer on". If it does, there's a problem. In fact, the switch on the PSU should NEVER be turned OFF unless you are going to be working inside the case or replacing hardware, or there is an electrical storm in the area (In which case it should ALSO be unplugged completely from the wall as switching it off will stop nothing if there is a strike close enough to be damaging) or when doing a hard reset procedure. Otherwise, there is zero reason for having the PSU switch off under any normal conditions. And, if you ever flip the switch on the PSU on and the system immediately tries to power on without using a power switch or jumping the pins on the motherboard, then there is a REAL problem with something because it should never do that.

So, I'm almost tempted to say there is an issue with the power switch or something else on the case. I'd recommend you disconnect ALL of the wiring and connections coming from the front of the case to the motherboard and try powering on the board using the pwr pins as explained here.

View: https://www.youtube.com/watch?v=FuPZlliGqBw

Oh ok. I will do that, when I'm back home. Can I damage something, if I bridge the wrong pins? Thank you so much, I learned a ton from you. On Saturday my parts for my gaming pc will arrive and I can also test if the PSU is the main issue.

 

[gosha1100]

Oh ok. I will do that, when I'm back home. Can I damage something, if I bridge the wrong pins? Thank you so much, I learned a ton from you. On Saturday my parts for my gaming pc will arrive and I can also test if the PSU is the main issue.

PSU switch should never "turn the computer on". If it does, there's a problem. In fact, the switch on the PSU should NEVER be turned OFF unless you are going to be working inside the case or replacing hardware, or there is an electrical storm in the area (In which case it should ALSO be unplugged completely from the wall as switching it off will stop nothing if there is a strike close enough to be damaging) or when doing a hard reset procedure. Otherwise, there is zero reason for having the PSU switch off under any normal conditions. And, if you ever flip the switch on the PSU on and the system immediately tries to power on without using a power switch or jumping the pins on the motherboard, then there is a REAL problem with something because it should never do that.

So, I'm almost tempted to say there is an issue with the power switch or something else on the case. I'd recommend you disconnect ALL of the wiring and connections coming from the front of the case to the motherboard and try powering on the board using the pwr pins as explained here.

View: https://www.youtube.com/watch?v=FuPZlliGqBw

Ok I got the Mainboard out of the case and it has actually a power on button on it.
I found out something important.
I got the black screen again and the mainboard showed b7 on the little screen.
I googled it and I found out that this means somethings wrong with memory and a DIMM crashed or something, but I‘m not sure.

 

[Darkbreeze]

The power on button is great and all, but I'd still try it using the pins.

 

[gosha1100]

The power on button is great and all, but I'd still try it using the pins.

Ok I will do that in the future, do you have any thoughts on the b7 error?

 

[gosha1100]

The power on button is great and all, but I'd still try it using the pins.

I plugged in my new PSU today and it seems to work now.

So the issues were a old PSU and a old CMOS battery.

Thank You for the help 😀