Well the NAS crashed again so not sure what to do next now. I guess I could go with plan B, remove the drives and test with a couple spare drives to see if it disappears again.
I was aware of the VGA header, would it be worth connecting it to a monitor? Does it show a real time log or something? It might help in troubleshooting what is going on.
I've noticed that if I unplug the network when it's not reachable, the front panel informs me the network has changed so that tells me the kernel is not gone.
I guess I shall try an OS reinstall after all. It can't hurt, right?
With ReadyNASOS installed, the only advantage of the VGA header is to see what's happening during boot, before you can start an SSH session, or to make BIOS setting changes. Once booted. it gives you nothing you don't get from SSH (unless, perhaps, SSH also crashes).
I'm no expert here, but that sounds like you have a "tainted" kernel. That the LCD driver is what called it is likely irrelevant except that it is called due to specific actions. The VPD process is also linked, as it lets the OS know which kind of display you have, but I think you'd see a much worse situation if the VPD file were corrupt (NAS would not boot).
There may be a way to fix (replace?) the kernel via support mode, but that's not in my wheelhouse. And I would think an OS re-install would already do that.
I was trying to work out a way to do a full OS re-install and retain your data by EXPORTing your volume, using one drive to factory default, importing the original volume (less one drive, so degraded), then DESTROYing the new one. I did so on a new-to-me RN3220, but the OS and swap partitions didn't propagate to a RAID including all drives, they remained only on the single factory-defaulted drive. I could manually add the missing ones to the OS RAID via SSH, but not to the swap. After the DESTROY, there was no longer any swap partition (but the OS expected one) and things got completely locked up. I had expected the DESTROY to create a swap partition on the remaining drives. I had to factory default the whole thing just to boot -- an OS re-install wasn't enough. I plan to try again, disabling the swap partition and trying to add the other drives before the DESTROY. But with your situation, doing that could make things worse. And I'm busy doing other things with that NAS now.
Will I be able to plug a keyboard and use the terminal via VGA? If so, why not.
For now, I've unplugged the LCD screen. I also never realised that the flash memory is at the front of the unit and if you unplug both those IDC cables, the NAS won't boot. 🙂
I've also stopped nv6lcd. It might be nonsense but my NAS has been doing this for years - only not so frequently. Can I assume the LCD is a USB device? What if it's faulty and it's crashing the system?
I'll attempt an OS reinstall - if someone could confirm whether the link I quoted above is the correct procedure as I saw another one online which suggests that holding the reset button for long results in a factory reset... 🙂
And no worries @Sandshark, I appreciate the thought. Unfortunately I am not incredibly skilled at Linux so I don't really want to try myself. If I had to factory reset the unit, I would look into an alternative software given that Netgear seems not to be very active with the development.
But I'd be very happy to just be able to use the NAS as it is 🙂
OK. I thought you'd already tried an OS re-install. Yes, that should have been the first thing to try. You will need to have the display re-connected when you do it, though. While units without a display use the system LEDs to let you choose menu items, those with a screen use only the screen, they don't also use the LEDs.
Don't worry about holding in the reset button too long. On a Pro6, that will make no difference. On units without a Backup button (rack-mount units), holding it longer makes it look for a USB recovery drive. But if one's not inserted, then that has no effect, except that it may have to be power cycled to boot (dunno, never tried it).
As @StephenB said, you can use a USB keyboard with the VGA display to access the console. But as I said before, it doesn't gain you a lot in your case. It's more important if the NAS isn't booting to check BIOS settings and see the boot progress.
If it crashes again I'll surely do an OS reinstall. I'm putting the VGA adaptor together, I know it's a shot in the dark but it might give us more clues as when the box disappears from the network, there is still some activity in it (the "network has changed" message when I plug/unplug the ethernet cable).
Finger crossed, I'll keep you posted. Thanks for now.
I know, it's just that removing my whole storage doesn't appeal me! 🙂 I'll see if I can do a partial backup so I can at least restore my document folder etc.
'cause yeah, the NAS has gone offline again, it only lasted a couple of hours. 😞
I know, it's just that removing my whole storage doesn't appeal me! 😞
I get that, but am thinking that you will either end up with factory default or a new NAS. Just guessing that a new PSU won't solve it, and you've basically exhausted all the other hardware options.
So one way or another, you need to get as much data as you can backed up.
Your problem doesn't seem like a power supply issue to me. Periodic voltage droops could cause that, but it's just not a common failure mode. A failing capacitor in the supply might do it, and "bulging" capacitors are a common failure on some supplies, but there are no instances of it on a Pro6 I can recall.
The original power supply is 300W. But if you look at the voltage specs, it has more 12V current available than a typical 300W. I recommend at least a 350W replacement, and I typically use a 400W. I generally buy ones that have automatic fan speed control, which tends to keep them more quiet. You can connect a standard ATX12V supply externally and see if that fixes the issue. But in your case, you're going to need to run it a while, so be sure to keep the airflow as intact as possible. Sealing the fan opening for the original supply and the gap where the cables come out with masking tape will accomplish that.
I am a bit into electronics, I've seen power supplies which look perfectly fine on a scope but after a while would start delivering unstable rails. That usually results in the unit locking though, not in this kind of partial crash. But when power delivery is not great, you never know, anything can happen.
Indeed before I replace the PSU I'd like to try an ATX power supply I have, I'll see if I can keep the airflow as stock.
And while testing an alternative PSU, I can put the Netgear PSU under a load and keep an eye on it with an oscilloscope.
It won't take long to be honest: the NAS has been crashing daily for a week now.
The existing PSU was "serviced" some months ago, I fitted a noctua fan and blew the dust off. I do not remember seeing bulged capacitors but this unit is now 15 years old approximately and it's been on 24/7 for many years so I feel there is a chance the PSU is failing.
The other weird thing is that I don't remember the NAS failing while in heavy use. As I said, the NAS is my multimedia centre - I keep a 1:1 ISO copy of my blu rays - so when I watch a movie the NAS is delivering steady 30/40Mbit for a couple of hours non stop. Now I am transferring my Documents folder on my desktop RAID and it's been working fine. I only find it disconnected when it's been idle for a while.
The other weird thing is that I don't remember the NAS failing while in heavy use.
There are some background processes still running, so that doesn't rule out a software issue. But it does make your case for the PSU being the culprit less likely.
Little "plot twist" - assuming I'm not just being optimistic, the NAS has been up and running for two days now, no issues.
What I did was to disconnect it from the network. How do I know it's still up and running? I use a laptop to connect to it every few hours to see if it still pings - the laptop is otherwise sleeping so the NIC is offline.
If it stays online, the next step would be to connect it to the second NIC I have on my main desktop - so I can access it but it's away from my main network and also from the internet.
I really don't have much stuff happening on that NAS, it mostly sits there idle. But I have a fixed IP address and the FTP is accessible from the internet via port forwarding.
I don't know, it wouldn't be the first time someone's hacked into my NAS 🙂
Any ideas?
Oh, if the NAS stays online, then I don't have a hardware issue! 🙂
NAS is on the second NIC of my Desktop PC. It's been up for three days so good news.
From where it sits, there is no access to the internet and even my PC does not really know much about it as it's on a different IP than before so all the mapped drives etc are disconnected.
Still, I noticed this:
When the NAS is connected to the network, doing nothing, the system temp rise 10C. When I unplug it, it cools down. System, not much the CPU. But the HDDs are really idle. The NAS is sitting on the floor.
Again, the NAS is on a different IP family so anything that is trying to connect to it is looking for the previous IP. And the NAS can't connect to the internet from where it is.
Could that just be the admin page I kept open? Still, that raises the temps by 10C? Maybe this is a red herring but it looks weird.
I did another test, I left it plugged in but without the UI open.
Network activity stops but the temperatures don't drop. See below, I logged out at 17.30, it's 18.40 now. You can see the activity stopping at 17.30 but temps stay higher.
I did another test, I left it plugged in but without the UI open.
Network activity stops but the temperatures don't drop. See below, I logged out at 17.30, it's 18.40 now. You can see the activity stopping at 17.30 but temps stay higher.
There is volume activity (operations per second) that coincides with the all of the rises in temperature. So there is some background activity happening.
and what could that be? As I said, the NAS is on a different IP address so all my network shares point to a different IP.
Is there a way to monitor what is transferring data/accessing the drives via terminal?
Edit: it doesn't seem to be that though. See attached: 17:30 is when I logged out, all activity stops. But temps stay high until I physically unplug at 18:30.
Do we know where the temperature sensor is? Could that just be the NIC IC warming up a bit because it's being used and the temp sensor happens to be nearby?
and what could that be? As I said, the NAS is on a different IP address so all my network shares point to a different IP.
Might not be linked to network activity at all. There are background processes that kick in from time to time. If you use indexing, that will get updated periodically, and of course AV (if enabled) will run scans from time to time. Snapshot processing, and other btrfs stuff might also be kicking in.
Do we know where the temperature sensor is? Could that just be the NIC IC warming up a bit because it's being used and the temp sensor happens to be nearby?
The one that's spiking appears to be the CPU temp, not the system temp. The CPU sensor is part of the chip. In many intel chips, it is not reliable at low temps - it's purpose to detect high ones. There are some old posts here from Netgear saying that they clamped low readings for some CPUs, so the spike might not be as big as it looks. But it's been quite a while (7-8 years ago), so I con't clearly remember the details.
