ReadyNAS RN4220 wont boot up after shutdown from WebUI

Eredar wrote:

I heard you can just swap drives between READYNAS products and they should work fine yeah?

Between some ReadyNAS models.  In this case, you'd need another OS-6 Rackmount.  There aren't any desktop models with 12 bays.

Eredar wrote:

I just am so confused why a simple reboot (done through the web UI!!!) would cause it to never boot up again

It sounds like a component failed on power-up, and that the NAS isn't passing POST.

I'm tagging Sandshark​ - he has a lot of experience with the rack mount systems (my own are all desktop models).

I agree that this sounds like it's failing POST and that a hardware failure has likely occurred.   I do not know the proper baud rate for the 4220 console connection.  When I had a 4220 to troubleshoot, I inserted a video card in place of the 10GbE card.

Unlike its predecessor, the 4200, the 4220 does not use a standard motherboard -- it's Netgear unique, so troubleshooting is difficult because Netgear has provided nothing to start from.  One possibility is the RAM.  It's standard DDR3 1600 ECC -- very common for servers of that generation and pretty cheap these days.  You can try with just a single RAM module in the white slot closest to the CPU and see if it then boots, trying each module.  The CPU is another possibility, though less likely.  A power supply is highly unlikely given the dual supplies -- it only needs one, the other is a backup.  But you can easily try each by itself. 

I don't remember the flash on that one being a DOM as on the later units, so not user-replaceable. But on most ReadyNAS, the flash is unpowered except for in boot, so failure is unlikely.  I don't know about the 4220.  But with the network LED flashing, maybe that is a sign it's looking for an OS.  Does your router show the unit is connected?

If your NAS is set up with multiple volumes and none exceeds 8 drives, then  a smaller unit is a possibility for data recovery.  You can insert the drives from each volume separately since the OS is duplicated on all drives.  There are a lot more options if none exceeds 6 drives. Otherwise, you'll need a 12-bay unit.  You may be hard pressed to find a 12-bay unit down under and shipping from the US would be a lot.  But perhaps you can find a 3220 (same mobo, different processor) or 4220 that somebody is willing to part out to reduce the shipping and risk that the problem is in the mobo.  But your best bet is likely a standard 12-bay server on which you can run Linux and use that for data recovery.

Hello,

Thank you both for your replies i really appreciate the help. 

Update based on testing:

I looked at the video card route but i am only used to PCIE slots (Zoomer, i know) and unsure what card would even fit on this custom MOBO?

The RAM is indeed 2x4GB DDR3 1600 ECC.

I have tried using just 1 stick in the white slot closest to CPU, and the behavior is different. The server will boot and  turn off within 5-10 seconds. After a few more seconds it will attempt to boot again by itself, repeating the same behavior indefinitely (boot loop). This happens with either stick of RAM in any slot solo. When both sticks are in any permutation it goes back to original behavior of just sitting there with a blinking power light.

Assume it might be a RAM issue but will take like a week to get DDR3 EEC (Never used EEC RAM before!). Assume any speed is OK or will the server reject anything higher? Trying to find something that fits those exact specs in Aus has been a bit annoying (its either 1600 or ECC never both, assume ECC is more important!)

PC3L looks like a low powered option of DDR3, would that help prevent this in future? Something like: Samsung 16GB (4 x 4GB) PC3L-10600R REG ECC RAM M393B5270DH0-YH9

or would it be rejected as its not the same as the shipped RAM?

After you confirm, I will get 4 new RAM sticks by Friday probably and check in next weekend with results, if it fails i will look into a new NAS as if its a capacitor or something that is blown i am unprepared and unwilling to start fault finding shorts and soldering on new components.

I have 2x Raid5 volumes, each being about 5*18TB drives. So im using 10 Drives total out of 12 on the NAS, with one being a Free Drive, and one slot not working (came that way, assume damaged connector, if you push it hard the green lights up but then drops when you stop pushing, not stressed about that too much as 10 is enough for me).

So i can just dump the 5 from the main volume into another OS6 NAS and it will work? The problem is i am like 75% sure what order i put them in but not 100% as the with the 1 missing slot being in the middle i either went up down or left right, any easy way to tell without bricking the drives? I do have an external SATA to USB cable + power i can use to read the drives manually on my PC easily.

Thank you both again for all the help!

The 4220 has a single PCI-e slot, normally containing the 10GbE NIC.  It's an 8x form factor, if I recall correctly, but that doesn't mean it's 8x functionally.  To support the NIC, it likely is 8x, but I went with a 1x video card to be safe.

Actually, each volume will work by itself in a 6-bay NAS.  It'll complain about the missing volume, but you can recover the data.  Do you have an old log .zip file you saved?  It will contain the information you need on drive order.  Just be aware that in the logs, the drives are numbered 1-12 in some and 0-11 in others.  If you boot in read-only mode, it won't do any damage if you have to experiment.  Just be sure to mark the drives as you remove them so you don't get everything mixed up.

If you want to keep using two 6-bay ReadyNAS, you can simply DESTROY the missing volume on the one containing the original main volume.  The other will be trickier due to apps (if you have any) and user private shares, which will be missing.