Disaster Strikes! Advice requested!

well I would remove power, carefully take out all 4 disks and label their order, test each on a PC with the disk manufacturers diagnostics tools (like WD tools, data life guard, etc).

while testing the existing disks, I would see if a new/used/different disk will boot up by itself, IE factory default with a separate disk, it won't affect anything on the 4 removed disks.

this will at least determine if the unit itself is failing. There are some known issues with certain devices that have power supply problems and these are covered under warranty. You would have to contact online support and provide them your device serial to see if yours falls under the known series.

However if the device powers up and appears to work with a new drive, then it doesn't sound like a power supply issue, and you may in fact have a dual drive failure.

unfortunately if thats the case, then likely your only solution would be starting from scratch with good drives and/or possibly sending the drives in to some kind of data recovery service.

Thanks for the response, the two disks that have no LED status do in fact test as bad. I don't think there is a power supply failure since there is power, 2 of 4 disk LEDs light up, etc. I'm kind of puzzled by the fact the LEDs are not lit, as opposed to blinking. I'm guessing I have to do a reset and start completely over, but I don't want to go thru the reset if there is some other problem or if there is some way to get the device to boot w/ new disks.

there is no way to recover from 2 disk failures other than reset with working disks, there is simply insuffucient data remaining on the 2 good disks to rebuild the 2 new disks.

if somehow you can get 1 of the bad disks working (meaning you have 3 original disks booting), you could theoretically add 1 new disk, let it resync get the array redundant again, then replace the other paritially/failed disk and let it resync.

else, you are pretty much screwed and learning the hard lesson that raid is not a substitute for backups (meaning multiple copies of your data, preferrably on different devices in different locations).

Thanks again for the response, I understand I can't rebuild the data on the disks, and I understand about backups, this device is a backup. I was just thinking it might be possible to somehow boot the os and generate a new raid set w/o destroying my os configuration in the nvram. I also don't want to go thru trying to start completely over if there is some additional problem w/ my hardware, which I sort of suspect since the blank LEDs are not what I expected for simply failed disks.

If a disk fails completely, the led will be blank. I have had this on two different disks. Normally, a disk will give you some indication of degradation of the platter and you will get warnings. However, in the case of a complete mechanical failure or a failure in the on board electronics, it will just die on the spot. You will see a similar reaction if you have a functional healthy array and are increasing the size of disks. When you hot remove the first one, the light will go out and it will be reported as dead.

When I lost my first one like that, it was overnight on a two drive array. I came in to the computer room to only one disk light. I shut the unit down normally and removed the first disk. It was cool to the touch. I then removed the other drive and it was warm to the touch. This confirmed to me that the first drive was indeed dead and I replaced it and rebooted the NAS. I was concerned because the drive that failed was only a month old. After the resync, they ran 24/365 for three more years before being replaced by larger drives.

zeke56 wrote:
I was just thinking it might be possible to somehow boot the os and generate a new raid set w/o destroying my os configuration in the nvram.

the os and config are stored on the disks, so if the array won't boot, then you have lost the config as well.

I also don't want to go thru trying to start completely over if there is some additional problem w/ my hardware, which I sort of suspect since the blank LEDs are not what I expected for simply failed disks.

well you can test hardware if you have a fresh/unused disk, just factory default with the one disk (doesn't take but a few minutes), then you can try booting it in each of the slots.

PapaBear wrote:

If a disk fails completely, the led will be blank. I have had this on two different disks. Normally, a disk will give you some indication of degradation of the platter and you will get warnings. However, in the case of a complete mechanical failure or a failure in the on board electronics, it will just die on the spot. You will see a similar reaction if you have a functional healthy array and are increasing the size of disks. When you hot remove the first one, the light will go out and it will be reported as dead.

Thanks, this is useful to know. To your knowledge is is possible to generate a new array using the installed firmware? The info on recovery I can find is geared toward rewriting a firmware image to a corrupted flash drive. As far as I know my firmware image is ok, I just have no working raid disks.

the device does not store any config, or run from the firmware.

when the disks are initialized, the firmware installs the os to the drives, so wherever/whatever the drives go, so does the config/os.

if you can not boot, you can not save your config, you will have to setup from scratch.

Confirmed. The copy of the OS on the firmware is there to provide a source for the OS when setting up a new array. The copy of the OS that they system operates from is on the drives as is the configuration. It is part of the array. When the array is lost, so is the configuration and the installed OS.

This means you must establish a new array, copying the OS from the flash memory, and establishing a new configuration. Unless you have a configuration that is very complex, it is not all that bad. We have actually spent more time discussing it than it should take to re-do it. The first time you go through it, it can be confusing, but once you have done it, it is easy to just follow the script and make each choice again.

Thanks both for the information. RE the comment about how long it will take to regenerate the configuration, believe me it was worth the effort to find out if it was possible to get around doing it. It was also a useful learning experience on my part. My big mistake was not making a backup of my configuration I guess. It didn't really dawn on me that the os actually only runs from the array. As I understand it the copy on the array is simply a mirror of the firmware + user configuration. This seems like a strange way for the device to operate, for reasons like this.