[SOLVED] RAID 1 - "degraded"??

[Squids4daddy]

Some background, then two questions. Background: My wife's computer(s) kept having hard drive failures at the worst possible moment, so many years ago I built a new machine where "C" was a RAID 1. I've noticed over the last 2 years a lot of failures. At one point, a lizard threw itself bodily across the RAID controller, so I replaced that. Until this year, two things have always been true.

First, more failures than I would like to see-maybe once a quarter. Second, those failures were always simple to deal with -- either a) just rebuild the array and that was good for 3 or 4 months or b) slot in a new drive and rebuild--good for 4 or 5 or 6 months. Now, sometimes "slot in a new drive" was ACTUALLY slot in an OLD drive. So let's say the original array was drives A and B and I slot in new drive "C" in place of "B". Well, 6 months down the road, maybe I shove "B" in to replace "C" and rebuild. That would succeed and be stable for 4, 5, 6 months. All necessary disclaimers about "yes I would take both drive C and B and plug them into a different machine for diagnostics to figure out that no, these drives had not really failed and were still good".

Since replacing the raid controller (the lizard suicide one) this year, I've noticed two (three) things. First, the array now "degrades" very frequently--every couple of weeks. Second, simply rebuilding the array or using an old hard drive, simply doesn't work. It gets to 1% and errors out. Now--this is important--if I take the old hard drive to a different machine and FORMAT it (NOT Quick format, old school full format)--it will rebuild with the "old" drive. But the array only lasts a couple of weeks. There is NO DIFFERENCE in array stability between a truly new fresh from the box HDD and "Formatted" drive. I get a couple of weeks, tops.

Specs: RAID Controller:
StarTech.com 4 Port PCI Express 2.0 SATA III 6Gbps RAID Controller Card with HyperDuo SSD Tiering - PCIe SATA 3 Controller Adapter (PEXSAT34RH)

HDD: WD Blue 2TB PC Hard Drive - 5400 RPM Class, SATA 6 Gb/s, 256 MB Cache, 3.5" - WD20EZAZ

MOBO: Asus Z87-A

Questions:

1. What could be causing the multi-year experience of having to rebuild degraded arrays (and what to do about it?)
2. Why the sudden inability to keep the array stable for longer than a couple of weeks, and the other aspects of having to fully reformat? Note that I replaced the "lizard suicide" controller because it started dropping the array every couple of weeks--the new RAID controller doesn't seem to have fixed that so maybe it's "something else"?
3. Is there a "better way"? Right now I do keep an updated daily Macrium image. But restoring from an image is a real pain the ass, it's slow, AND not as reliable: it's the "please dear God let this work" of last resort. The advantage of RAID 1 is that if a HDD fails, the wife can keep working until bedtime at which point I can rebuild the array "overnight" and know that it will be fine in the morning.

 

[mdd1963]

I'd just extricate what files you need to an external drive, abandon RAID 1 and get a decent SSD/fresh install, and, confine the spinning drives to offline backup images...and/or periodic/incremental backups... Store important stuff on the cloud periodically.

It'd be hard to know what the RAID issue is/was, be it marginal drives actually failing, or, errors induced due to RAID controller, RAM, who knows.... (most folks don't have RAID 1 issues every two-four weeks, so, I can understand your frustration with it)

 

[USAFRet]

Q1: One of the drives has issues, and they become unsynced.

Q3: I use Macrium in my daily regimen, and have never had problems. Not a one. I suspect the RAID situation and a possibly errant drive or controller is the issue with recoveries.

 

[popatim]

I'm thinking its the firmware of the raid card. Some of those were pretty buggy but to upgrade that, if the files can even be still found, we'd need to know the revision of the controller chip to locate the relevant firmware.

I agree with give up on raid 1 if you can afford the pc being down long enough to restore from a backup.

 

[fzabkar]

Have you examined the logs to see which sector on the failed drive was responsible? You could then use a disc editor to examine this sector on each drive.