Question about TLER

Question

Level 4

2,096 points

Question about TLER

I have reconfigured my HD configuration and just started using a Western Digital 400 Gb RE2 as my non-raid boot drive. This drive has TLER (Time Limited Error Reporting) as standard.

I aware that the official Western Digital advice is not to use TLER disks in non-raid mode. The reason given is that when an error is detected the TLER will only allow it to try and recover for 7 secs (the logic being that raid would take care of the problem), while a non-TLER disk will go on trying for longer and hence presumably be more likely to successfully remap the sector and recover.

My way of looking at it is that if errors are happening I want to know about it. Once a disk starts having errors I want to replace it. A non-TLER disk will conceal the problem for me for longer than a TLER one. All my data is backed up twice so I can recover easily from a failure.

Am I missing something? How often does the error detection/remapping process occurr on a modern disk?

Thanks for any input.

MacPro 2.66(2006), 8 Gb, 4x1Tb, HD3870, MBP 1.83, 2Gb, 500 Gb,, Mac OS X (10.5.6), 30" ACD, Sonnet eSATA, LG Blu-Ray+112D. XP Parallels, Vista Boot Camp, iPhone 3G

Posted on Apr 21, 2009 5:29 AM

Reply

Answer 1

The hatter

Level 9

61,031 points

Apr 21, 2009 6:24 AM in response to Mike Boreham

We use to kick this subject around a lot when the 500GB RE came out (I wasn't even aware of a 400GB RE2). And at some point I thought (maybe wrongly) that WD softened up and that use in non-RAID (and not on Hardware controller) was acceptable. Using drives not optimized for RAID was more "frowned" on.

There are SSDs of course, trim down your OS, or there is the 10K VelociRaptor, too, or even the Caviar and Caviar Black 640's which sell for ~$77 and nice, fast drives for various uses.

*Q: What is time-limited error recovery and why do I need it?*
A: Desktop drives are designed to protect and recover data, at times pausing for as much as a few minutes to make sure that data is recovered. Inside a RAID system, where the RAID controller handles error recovery, the drive needn't pause for extended periods to recover data. In fact, heroic error recovery attempts can cause a RAID system to drop a drive out of the array. WD RE2 is engineered to prevent hard drive error recovery fallout by limiting the drive's error recovery time. With error recovery factory set to seven seconds, the drive has time to attempt a recovery, allow the RAID controller to log the error, and still stay online.

http://www.wdc.com/en/products/products.asp?driveid=400

I would not worry about it, and if you want actual I/O error reporting, buy and use SoftRAID to format your drives.

Reply

Answer 2

Mike Boreham Author

Level 4

2,096 points

Apr 21, 2009 6:48 AM in response to The hatter

Thanks!

Interesting you should mention SoftRAID which I already have. My Mac Pro has five drives in it, the boot WD400RE2 in the lower optical bay, and four 1Tb drives in the trays. The four 1Tb are all converted to SoftRAID and have raid volumes made with SoftRAID.

At this point I have not converted the WD400RE2 boot drive to SoftRAID format, because I didn't see a reason to, and because I am a bit scared of converting the boot drive to a third party format. Do I need to be scared? I know some utilities don't work with SoftRAID drives, although I believe Diskwarrior does. Any other issues with converting to SoftRAID?

I know there are a lot faster drives than the WD400RE2 for boot but I have had this in a drawer for a year or two, and my testing shows that faster boot drives don't give much benefit except boot time and prog launching. The scratch four disk striped raid (from outer 20Gb of each 1Tb) makes CS4 really fly. Also 400 Gb is nice size for a boot drive. I'll probably skip the Velociraptors at this stage and wait for SSD to come down a bit in price.

Reply

Answer 3

The hatter

Level 9

61,031 points

Apr 21, 2009 7:17 AM in response to Mike Boreham

I never ran into a program that balked at SoftRAID. Ask their support team, news to me but possible. They pound drives like you and I couldn't if we tried in a dozen years, and the convert to/from Apple partition - more for RAID but also for single drive - has been around for years. They added extra backup security, too.

There are some cosmetic issues with GPT and EFI 'volumes' showing in some utilities, but other than that....

I rotate boot drives every year or two, and keep older drives for an emergency backup only. Or archive of older stuff off line.

I probably would have opted for 80GB x 4 or larger for scratch, and 2-3x what gets used during a run. I've seen CS eat up quite a bit of space even with 500MB file images.

Heck, the WD Caviar Black hits 100-120MB/sec for the outer 1/5th and using just 200GB? for system would be nice. RE2 was more like 75MB/sec as I recall in my benchmark file.

Reply

Answer 4

Mike Boreham Author

Level 4

2,096 points

Apr 21, 2009 7:31 AM in response to The hatter

Thanks for all that.

Since the conversion to SoftRAID is easily reversible I'll give it a try, and take advantage of SoftRAID's error reporting.

Reply

Answer 5

Mike Boreham Author

Level 4

2,096 points

May 4, 2009 8:16 AM in response to Mike Boreham

Well, been using SoftRAID for two weeks now and no problems until today. Suddenly had a spinning rainbow for a few seconds, followed immediately by a SoftRAID error warning that my non-RAID boot volume (the WD400RE2 with TLER) had an i/o error.

No other symptoms and it has been OK since.

I had SoftRAID monitor set to show warning for recoverable and non-recoverable disk errors (default is only non-recoverable). Seems like mine was recoverable (since it recovered)

Is it time to junk the WD400RE2?
How often should I expect to get recoverable disk errors?

Thanks for any advice.

Reply

Answer 6

Mike Boreham Author

Level 4

2,096 points

May 4, 2009 8:24 AM in response to Mike Boreham

Duplicate post

Message was edited by: Mike Boreham

Reply

Answer 7

The hatter

Level 9

61,031 points

May 4, 2009 9:19 AM in response to Mike Boreham

WD4001ABYS is the current 400GB RE2.
http://www.wdc.com/en/products/products.asp?driveid=400

I think it is helpful (or essential) anymore to zero a drive - with WD Diagnostic utility in Windows being best - before using.

Keep your eye on it, it you get two non-recoverable, then if you have Windows, definitely take steps. Otherwise, you can try using a full zero all of the drive (helps if you know where to "zero in on" to focus, so you could zero or even 7-way write).

SoftRAID will let you do a zero all, and you can cancel at any time, but if there are bad sectors, not always map those out successfully. And you can see how many by looking at system.log when you click on Disk Warrior's manual diagnostic test.

One note, too many cooks spoil the soup, and get in each other's way.
Having more than one tool that is testing and monitoring drive behavior can cause problems, even performance or as far as stability. So if you use SoftRAID, don't also use Hardware Monitor to also check your hard drives. Less is more and minimize all such "SMART" tests etc.

Reply

Answer 8

Mike Boreham Author

Level 4

2,096 points

May 4, 2009 9:31 AM in response to The hatter

The hatter wrote:

One note, too many cooks spoil the soup, and get in each other's way.
Having more than one tool that is testing and monitoring drive behavior can cause problems, even performance or as far as stability. So if you use SoftRAID, don't also use Hardware Monitor to also check your hard drives. Less is more and minimize all such "SMART" tests etc.

Thanks for reply!

Intrigued by your comment above, which makes sense. I don't have anything else SMART monitoring the HD condition, only SoftRAID. However I do sometimes leave Hardware Monitor running. I use it mostly for temps. I don't think it does any SMART type condition monitoring. Are you saying that the temp monitoring "could get in the way" of SoftRAID monitoring?

Reply

Answer 9

The hatter

Level 9

61,031 points

May 4, 2009 9:59 AM in response to Mike Boreham

Turn off HM's drive monitoring for when you use it, it will check/probe the drive's SMART status every nn seconds or minutes.

Even Disk Warrior seems to have option to automatically check drive "integrity" or status I think.

Intech Speedtools has the most thorough SMART test which can take 45 minutes but is run manually.

Mike Breeden had some notes about variations on SMART results from vendor's own (accurate) and how there is no one standard or utility that is accurate.

Reply

Answer 10

Mike Boreham Author

Level 4

2,096 points

May 4, 2009 10:38 AM in response to The hatter

To underline the advice about too many cooks....

I just got out a spare WD7500AAKS drive to replace the WD4000RE2. It was converted to Softraid and I started surface scanning it with Drive Genius. Within about two minutes I had four SoftRAID errors. Now converted it back to Apple and scanning again.

Do you think the original SoftRAID error on my WD 400RE2 could have been due to Hardware Monitor running, not a genuine error at all?

Reply

Answer 11

The hatter

Level 9

61,031 points

May 4, 2009 11:53 AM in response to Mike Boreham

I would only be guessing what happened, why, what was actually happening at the time.

TechTool Pro does a fairly good, reliable surface scan, one of its many features.

Disk drives today really should be 'inspected' before using.

If it isn't part of a RAID....

Just pulled my OEM Apple RAM, and my system is better, memory temps are down - and I thought it was the other RAM causing the spike in temp. I think some problems were from mix of RAM. I say that because there are a lot of things to account for in system stability.

You might be better off with another drive, say WD Caviar Black (AFTER zeroing etc).

RE2 400GB, was that bought with a set? to be spare of a RAID array? that would be the only time it would make sense.

TLER was heavily discussed when it first appeared in WD specs. And there was a 400GB model that had SSC (spread spectrum) that would not work in G5s.

An error is an error, real and geniune.

Shoot an email over to SR support about running HM etc or TLER.

Reply

Answer 12

Mike Boreham Author

Level 4

2,096 points

May 4, 2009 12:15 PM in response to The hatter

The WD400RE2 started out as one of a raid pair in a Datawhale eSATA box. I can't remember the details now but the other one started getting SoftRAID errors so I quarantined it (should have pursued warranty, but it passed every other test I could do on it), so in the end I gave it to a friend, with it's history, and has been OK for him for 18 months or so. So I just have the one now.

I have been in touch with SoftRAID support. They tend to discount (but not completely) the possiblity that Hardware Monitor, SMART or Drive Genius are interacting with SoftRAID's monitor.

They emphasize the OSX suppresses up to 75% of errors, that SoftRAID reports.

BTW how do you turn off SMART in Hardware Monitor? can't see that you can.

Reply

Answer 13

The hatter

Level 9

61,031 points

May 4, 2009 1:29 PM in response to Mike Boreham

Don't monitor that or any drive in HM. There is no "SMART" disk feature - or there are but "RAID Status" and other items. And there is a footnote about not testing a disk "frequently."

I'd pull the drive given the history until you can zero it with WD Lifeguard Diagnostic.

OS X does "skate on thin" ice. Too often and too much so.

Reply

Answer 14

Mike Boreham Author

Level 4

2,096 points

May 4, 2009 2:16 PM in response to The hatter

The hatter wrote:
Don't monitor that or any drive in HM. There is no "SMART" disk feature - or there are but "RAID Status" and other items. And there is a footnote about not testing a disk "frequently."

Understand now. I have had my Smart disk interval on 1 hr.

I'd pull the drive given the history until you can zero it with WD Lifeguard Diagnostic.

I will try and do this in Bootcamp (on my laptop) via SATAto USB bridge or eSATA PMCIA 34 thingy. have to get the WD7500AAKS installed as replacement boot drive first.

OS X does "skate on thin" ice. Too often and too much so.

Maybe but I have a learning curve to understand how to react to all these SoftRAID errors!

Reply

Answer 15

Mike Boreham Author

Level 4

2,096 points

May 5, 2009 1:27 AM in response to Mike Boreham

Mike Boreham wrote:

The hatter wrote:

I'd pull the drive given the history until you can zero it with WD Lifeguard Diagnostic.

I will try and do this in Bootcamp (on my laptop) via SATAto USB bridge or eSATA PMCIA 34 thingy. have to get the WD7500AAKS installed as replacement boot drive first.

Interesting. I have put the WD7500AAKS in the Mac Pro as boot drive and all seems good there. Have been trying to investigate the 400RE2 with WD Diagnostic Tools with Vista Bootcamp on my MBP.

It connects fine with the SATA express card, and WD Diagnostics sees it OK, but always reports "test cable error" immediately if I try to run the tests. At first I thought this was a problem of the set up, but another WD drive (old Raptor 74) tests fine in the identical set up with no Test cable error. Also tried a different SATA cable and a two different USB set ups with same results.

Perhaps the test cable error is a problem with the HD.

Message was edited by: Mike Boreham

Reply