2x 2TB WD green in ReadyNAS Duo?

I see this: http://www.readynas.com/?p=3690

Your referenced article is over a year old. In that time WD and the ReadyNAS line have made significant changes.

The best way to determine the compatibility of a drive is to choose one that is on the Hardware Compatibility List [urlhttp://www.readynas.com/?page_id=82](HCL)[/url], which lists drives that have passed the testing of the ReadyNAS staff.

I would like to repeat a question that has been asked repeatedly about WD Green drives on the HCL, which as far as I can tell has not been officially answered by Netgear or the Jedi. This posting is inspired by the fact that I just had to replace a 1 year old 3 TB WD Green WD30EZRS which is on the HCL.

I installed 3 of these disks on my Ultra 6 Plus when I bought the unit a year ago. Since the disks are on the HCL, and since I didn't have a convenient PC with a SATA port, I ignored the LCC issue. After 3 or 4 months, all 3 drives showed LCC's of over 250,000 in the SMART status. Since WD states they only test the drives to LCC's of a million, I bit the bullet, put the disks in a suitable computer (a Mac Pro), and ran wdidle3 on them; this stopped the LCC's from increasing. All was well until a few days ago, when my readynas went totally dead (which I described elsewhere on the forum); after I resurrected it (with much appreciated help from Jedi Knight and other Wise Ones available on this forum) I found that one of the WD drives had increasing write errors in the SMART status. The readynas did not consider the drive bad, but I replaced it with a new Seagate drive (also on the HCL).

So.....can we safely use WD drives on the readynas? - the drives are on the HCL - and if so can we ignore the LCC, or do we need to have the means and expertise to run wdidle3. And, given that the drives are on the HCL, can netgear provide a more convenient tool than wdidle3, that will run natively on the readynas - either built in or an add-on? I have never enabled root SSH, for security concerns, but I can if I need to.

Thanks!

While I cannot offer any opinion on the LCC and wdidle3 (I do not use WD drives in my NAS units), keep in mind, any drive can and will fail, usually at the most inconvenient time. A give drive may fail early, while it's neighbor in the NAS may run for years. When I first installed my NV+ almost 4 1/2 years ago, I installed 2x500GB Seagate drives. Within a month, one was dead. When I pulled it out it was cold to the touch. I replaced it with another of the same model and the unit resynched. I then ran the unit for almost 3 years, slowing adding more drives (different manufacturers and sizes). When I upgraded the drives to 1TB, the one remaining original drive had over 25,000 hours on it, and the replacement for the one failed drive almost as much. Seagate replaced the failed drive.

A year and a half ago, I installed two NVX units, each with 4x1TB Seagate drives (4 pulled from the NV+) and 2 spares purchased at the same time. Out of the 10 drives, in the year following, I have lost two more, both of which have been replaced by Seagate.

If you are having write errors, I don't think that is related to the LCC. While one can purchase Enterprise class drives which will reduce the chance of failure, it will not eliminate it althogether. After all you have electrical circuitry on a drive and mechanical movements within a drive. Either can wear and fail. This is why one should never trust their data to a single device, even if the device is an NAS with redundancy. After all, the failure could occur within the NAS, or one could encounter a catastrophic loss of the unit itself. If one reads through the entire manual, they will be reminded that the NAS is not a backup plan.

You're correct about backups and disk reliability, of course. Which is why (per your earlier advice, BTW) I back up the important data on my readynas ultra plus to my old, but still functional, NV+. I have a USB backup as well. But even with backups disk failures are a pain (or worse) and the fewer episodes one has to deal with the better. My current troubles did in fact include complete failure of my readynas, which, fortunately, was reversible.

There really is an issue with the LCC which Netgear needs to address. Several WD "green" drives (the 3TB WD30EZRS and the 2TB WD20EARS drives, for example) are on the official Readynas HCL and have the LCC issue.

WD does not endorse these drives for use in a NAS Raid device. The major issue seems to be the LCC. Out of the box, these drives will run up a LCC of over a million in less than a year when installed in a Readynas. WD states that the drives have only been tested with LCC's of a million, over this is uncharted territory.

The way to fix this, if indeed it needs fixing, is to mount the drive on the internal SATA port of a computer that can boot DOS, and run the WD utililty wdidle3. But this is inconvenient, and may be beyond the technical expertise of many users. And, of course, it requires a suitable PC.

So, since the drives are on the HCL, I think Netgear needs to address whether the high LCC count is likely to reduce the drive lifespan, and if it is, I think they should try to prepare a readynas based solution that does not require mounting the drive in a PC and running wdidle3 (which requires booting in to DOS)

Actually, it is not just the "green" drives that WD recommends not be used in RAID service, it is the entire WD consumer line of drives. WD recommends on the Enterprise class of their drives be used in RAID service.

And yet the only 3TB WD drive on the HCL is the consumer WD30EZRS.

A dilemma.

Personally, I think we have to follow the readynas HCL.

Oh netgear - any comments on the LCC issue?

Have you considered the Hitachi HDS5C3030ALA630? I have had four of these in service for about six months without any problems. In addition, right now they are less expensive than the WD30EZRS.

Good suggestion - you probably have the same Fry's ad I do!

A year ago, when I bought my WD30EZRS they were advertised as 7200 RPM drives, the hitachi model you mention was advertised as 5400 RPM drives. So, I went for the WD's.

But, probably, I was fooled. Both drives are "enviro" drives with variable spin speeds, the Hitachi's call this "coolspin" and the WD's "intellipower". At least on Tom's Hardware, the two drives have generally similar specs. And, the marketing for the WD30EZRS drives now calls them 5400 RPM drives as well. (WD's tech support tells me the drive speed varies between 5400 and 7200). (as another aside, the WD30EZRS has been replaced by a WD30EZRX model which has a SATA 3 interface. I think my readynas is SATA 2, so this probably is irrelevant).

An array is probably only as fast as it's slowest drive, so I wouldn't want to degrade a 7200-rpm array to 5400. But, in doing a bit more homework, it seems the Hitachi 5K3000 models you suggest is probably similar to the WD's I already have, so I may well expand with the Hitachi model as you suggest

The LCC issue, with the WD drives, as you can probably gather, drove me nuts.

Have you checked the SMART status with your Hitachi drives? Is the LCC increasing towards infinity and beyond as it does with a non-wdidle3'ed WD drive?

Thanks again!

The closest Fry's to me is over three hours away (Atlanta). Four of my five Hitachis came from Newegg and one from Amazon. In each NVX I have the 2x3TB Hitachi (5400 RPM) and 2x1TB Seagates (7200 RPM). I have noticed no apparent slowdown in the file transfer rate since I replaced two of the Seagates (originally 4x1TB) in each unit. The fifth Hitachi is a spare, so far unused and still in it's sealed anti static bag. Keep in mind that even the SATA1 drive interface (1500 Mb/s) is faster than the gigabit network rate (1000 Mb/s) so I would not expect to see much impact.

The two 5400 RPM SATA3 (6Gb/s) Hitachis seem to coexist well with the 7200 RPM SATA2 (3Gb/s) Seagates. I will be honest that I had not checked the SMART logs and when I did, the Hitachis did have an LCC number, 10 on one and 13 on the other, both of which matched the power off retract count. This was NAS1, so I checked NAS2, my backup target. The Hitachi in bay 1 has an LCC number of 304 and the one in bay 2 is 298. Again, both numbers match perfectly the power off retract count. The drives in NAS2 are set to spin down after 90 minutes of inactivity, NAS1 is not set to spin down. Keep in mind that different manufacturers use some of the SMART registers for different things. I don't believe the Seagates for example use the LCC counter, as I have never had one show up on a Seagate.

The two Hitachis in NAS1 have almost reached the 4,000 hour mark while the two in NAS2 have almost reached 3,900. I put the two in NAS1 first and about 5 days later repeated the process with NAS2. They have been in service now for almost 6 months, whereas the Seagates have been in service for about 18 months and have around 11,000 hours on most of them. (Started with 8 with 2 spares which have been put in service. They have also been moved around a bit with 7 currently in three different boxes (3 moved into NAS3 a 4 1/2 year old NV+ after being replaced by the Hitachis in NAS1 and NAS2). The NV+ was going to be rehomed after I got the second NVX unit, but that didn't happen so I put it to use as a secondary backup. [One can never have too many backups].