Help swapping "dead" drive? [Case #15693780]

mdgm wrote:
Did you run the memory test (http://www.readynas.com/forum/faq.php#Is_there_a_way_I_can_verify_if_my_memory_is_good%3F) twice after upgrading the memory? Memory upgrades are not supported and incompatible memory can cause problems.

I did not. So the device has a list of compatible disks AND memory chips? I assume you mean beyond the memory type/speed? I picked up the RAM from newegg when I bought the NAS. I actually got everything in one shot from them on the same day. Starting to regret thinking this device would "work" without all these potential fail points. Do folks have better luck with the more expensive "pro" or business class NAS units?

I accept all the blame for not RTFM before diving in to this, of course. Still frustrating!

Running the mem diag now. How long should that take? I will run it twice. I am prob going to head to bed shortly if it will take longer than 30 minutes. Thanks for the suggestion...is it possible that bad memory would cause my "missing" disk? Hmm

Memory upgrades are not supported? So they would expect us to stick with the installed 256MB? Yikes. It's 2011. :)

The compatibility list is here: http://www.readynas.com/hcl

Unless you bought the official NetGear 1GB memory module it's not supported.

The more expensive units also don't have supported memory upgrades (the desktop x86 [Intel] home/business units come with 1GB RAM) and have a hard disk compatibility list.

I haven't needed to run the memory test, so not sure how long it would take. Whether or not bad memory caused your problem you should run the test on the memory module and even if the memory is fine (passing the memory test does not guarantee this) tech support would want you to install the stock memory module.

The ReadyNAS runs heavily customised Debian Linux. It's not bloated like Windows or Mac. Take a look at the size of the firmware updates. The firmware updates contain the entire OS. For the NV+, 256MB RAM is plenty. If you want to run lots of addons or other resource hungry tasks you should be looking at an x86 (Intel) ReadyNAS anyway.

First memory test ran OK returning "Memory OK" with all LED's lit as expected. I will wait for it to boot up now, power it down and set it to test a second time.

I have the stock memory and can replace it if necessary. I did the upgrade to improve the Frontview performance. It's still pretty sluggish moving from tab to tab.

It was about an hour to run the memory test. Not too bad. I think I will call the support line tomorrow as I am still in the 3 month window for free phone support.

Thanks again for your assistance. Time for bed. Back to work tomorrow after a long weekend.

Second test ran OK as well. "Memory OK" - next?

The memory may be O.K., but it may not be.

The mention of an invalid RAID superblock doesn't sound good.

Due to using incompatible drives and a 3rd party memory module I would expect support to refuse to help. They can't be expected to support issues arising possibly due to the use of incompatible hardware.

You could try a data recovery company or attempt to fix the problem yourself.

mdgm wrote:
The memory may be O.K., but it may not be. The mention of an invalid RAID superblock doesn't sound good.

Can you be more specific on the memory? If the test runs OK, it can still be the problem? What does that say about the test?

mdgm wrote:
Due to using incompatible drives and a 3rd party memory module I would expect support to refuse to help. They can't be expected to support issues arising possibly due to the use of incompatible hardware.

Well I guess I know why the WD20EADS drives were $99 a piece on newegg. Only 2 months later, they are all mysteriously "out of stock" too. I wonder why. :roll:

So the "support" policy is not to help anyone who may not have read all of the documentation before purchasing their hardware? That seems like a slap in the face, really. I understand them not wanting to support issues they have no solution for, but there should probably be a bold faced warning on these devices that alerts consumers to the HCL requirements. Sucks to find out you've got incompatible hardware AFTER you've purchased it and loaded your personal data to it.

If the device has update-able firmware, why does it not alert us that the hardware we're using is not compatible? Maybe an idea for future releases.

Still my fault, so i will be stuck trying to fix it myself i guess. Maybe they'll surprise me? ha, I will give the TLER tool a shot tonight I suppose. Is XRAID a netgear proprietary format? Can I try putting the disks in another device to see if the data shows up? What a nightmare. :?

So I had another look at the logs just now, and I notice it claims the hdi [drive #4 which reports normal in all interfaces] disk has an "unknown partition table" - this drive is marked in the log as the Parity drive. Does that mean the "unknown" partition table is expected? Just for kicks, I ran the SMART test on this drive from Frontview:

SMART Information for Disk 4

Model: 	WDC WD20EADS-00W4B0
Serial: 	WD-WCAVY6067131
Firmware: 	01.00A01

SMART Attribute
	
Raw Read Error Rate	0
Spin Up Time	10150
Start Stop Count	13
Reallocated Sector Count	0
Seek Error Rate	0
Power On Hours	1487
Spin Retry Count	0
Calibration Retry Count	0
Power Cycle Count	12
Power-Off Retract Count	11
Load Cycle Count	70418
Temperature Celsius	33
Reallocated Event Count	0
Current Pending Sector	0
Offline Uncorrectable	0
UDMA CRC Error Count	0
Multi Zone Error Rate	0
	
ATA Error Count	    0

Anything look alarming here? LCC high?

Also tried a zero fill and full erase on my replacement disk before attempting to hot add it again, and I don't think it's an issue with the disks at all. Your talk about compatibility is worrying me. If the device built a RAID filesystem on these disks, how did they get to a place where they are now unrecognizable by the same device? Also, the fact that none of the disks i've tried [3 now counting the original "failed" disk, the other 2 are brand new and the one in there now is a WD20EARS even] show anything but "dead" or "not present" to the NAS.

@mjw3786
LLCs are high. The highest on my 6 drives is 2467 and all of the others are under 1000. I would definitely see if you can use both WDIDLE and TLER on your drives. My last drive was the one that had a different firmware (see Sig) and it was not supposed to be able to be "TLERed" but it was. When I did it stopped the LLC escalation in its tracks. I would try both tools and also replace the original memory to eliminate any variables.
Hope this helps. Please post results.
bigbearf

Well, i'm not happy with the result but it's better than losing the data. I will report back when I have had a chance to follow through on this, but here is what i'm left to:

Buy 4 "compatible" drives. Apparently, the reason they only support certain drives relates to a resync command that the compatible drives have and others do not. This causes the NAS to report drives as dead even though they're fine. Now for the precarious steps:

CLONE these 4 disks using ghost or something similar to the 4 compatible disks. Insert the compatible clones back in to my NAS and pray all is well. Ugh. Buying 4 more disks to test this out does not sound fun, but if it will rescue my data, it is worth the price of admission.

The tech was confident this would work. Thoughts on this before I shell out for 4 new disks on the HCL? I am going for these:

WD Caviar Green WD20EARS 2TB -- Firmware 80.00A80, Model # WD20EARS-00S8B1, See Tech Bulletin http://www.readynas.com/?p=3690

My only issue appears to be the -00S8B1 - How can I track down this specific model number? I have 2 brand new WD20EARS-00MVWB0 disks here now. One was sent as the RMA replacement from WD and the other I purchased as a hot spare from newegg. Can I use these? I had one in slot 3 when he logged in to my machine to look at the logs and said that it was on the HCL. So i'm slightly confused.

Where can I grab some of the disks on the HCL? Do I need to go right to WD for that?

BigBearf wrote:

@mjw3786 LLCs are high. The highest on my 6 drives is 2467 and all of the others are under 1000. I would definitely see if you can use both WDIDLE and TLER on your drives. My last drive was the one that had a different firmware (see Sig) and it was not supposed to be able to be "TLERed" but it was. When I did it stopped the LLC escalation in its tracks. I would try both tools and also replace the original memory to eliminate any variables. Hope this helps. Please post results. bigbearf

hey BigBearf, thanks for the continued info. the tech seemed confident my drives don't support the resync command, but I will give the TLER a shot on the "dead" drive and see how we do. I will report back when I have a chance.

running WDTLER [TLER-ON] yields disappointing [i think] results:

Unable to set Read TLER time
Error enabling Read TLER to 7.000 seconds.
Unable to set Write TLER time
Error enabling Write TLER to 7.000 seconds.
Unable to retrieve Read TLER time
Unable to retrieve Write TLER time

What should I be seeing here? Should I assume TLER is not supported by my disks? other options? I think I am going to look in to getting some WD20EARS disks to clone these expensive paper weights. I suppose I can use them as backup disks [not in the NAS of course]