How long will SpinRite continue to try

[markav2]

How long will SpinRite continue to attempt to recover? The photograph below was taken Monday morning, on Saturday morning it was only one block less than it is today.

It has now been running for well over 80 hours on a 320 GB hard drive.

When does it decide that it just can't recovery and to move on?

 

[ColbyBouma]

How long will SpinRite continue to attempt to recover?

DynaStat will run for up to 5 minutes, as show by the "Recovery" timer at the top.

it was only one block less than it is today.

1 block in the GSD represents many sectors. If you scroll a couple screens to the right and look through the log, you'll probably see a LOT of entries.

 

[ShadowMeow]

Have you tried RC5?

 

[PHolder]

There is a command line option to ask SpinRite to take less time on DynaStat. This can save time, but could possible miss a recovery if it needed that much extra effort and you limited it. You can even set it to zero, which will NOT do recovery and you WILL lose data. Ask for help on the command line for more details.

 

[ShadowMeow]

By running it on level 3 you effectively only had one chance to recover the data in it's original state. Level 2 would of been a better choice. Level 3 reads writes and modifies each sector taking away it's original state. It's good to know how to use the tool correctly.

Anyhow Good Luck!

 

[DarkwinX]

By running it on level 3 you effectively only had one chance to recover the data in it's original state. Level 2 would of been a better choice. Level 3 reads writes and modifies each sector taking away it's original state.

What exactly do you mean by this?

The only difference I know of between a level 2 and level 3 is that a level 2 is a selective rewrite and a level 3 is a total rewrite (with a level 4 being the same with a double rewrite using inverted data).

In 2 and 3 I don't believe dynastat does anything different with the recovery attempts. It tries very hard to determine what the data should've been from the block and then tries to rewrite it.

 

[ColbyBouma]

(with a level 4 being the same with a double rewrite using inverted data)

Level 4 was changed to be Level 3 + an additional read. Level 5 is basically what Level 4 used to be.

SpinRite 6.0's levels:

Level	Actions	Data Recovery
1	Read	No
2	Read	Yes
3	Read, Write	Yes
4	Read, Write inverted, Read, Write original	Yes

SpinRite 6.1's levels:

Level	Actions	Data Recovery
1	Read	No
2	Read	Yes
3	Read, Write	Yes
4	Read, Write, Read	Yes
5	Read, Write inverted, Read, Write original, Read	Yes

 

[ShadowMeow]

What exactly do you mean by this?

What I'm trying to say is...

Level 2: Leaves a sector as-is if SR cannot recover anything at all. Assuming "dynastat 0" isn't being used. Giving you a second chance with other tools like the RapidSpar Disk Imager or the PC-3000 Disk Imager.

Level 3: Refreshes the sectors, so it's more of a data refresher, but it writes back to every sector on the drive.

 

[ColbyBouma]

Level 2: Leaves a sector as-is if SR cannot recover anything at all. Assuming "dynastat 0" isn't being used. Giving you a second chance with other tools like the RapidSpar Disk Imager or the PC-3000 Disk Imager.

That is incorrect. DynaStat behaves the same on Levels 2 through 5. The levels dictate what SpinRite does to sectors outside of DynaStat, and whether or not DynaStat is engaged.

If DynaStat can't recover any data by the time its timer runs out, I believe it fills the sector with 0's. If it recovers partial data, it will write that back to the sector.

 

[markav2]

I have a very sick drive

 

[ShadowMeow]

That is incorrect. DynaStat behaves the same on Levels 2 through 5. The levels dictate what SpinRite does to sectors outside of DynaStat, and whether or not DynaStat is engaged.

If DynaStat can't recover any data by the time its timer runs out, I believe it fills the sector with 0's. If it recovers partial data, it will write that back to the sector.

OK. I stand corrected.

 

[ShadowMeow]

I have a very sick drive

View attachment 890

Repower drive.

 

[Steve]

Repower drive.

Just a note, James, that there are two very similar looking screens. One explains that the drive has declared a "FAULT" (in itself) and that power cycling of the drive will be required. The other one (shown by this user above) is a lesser problem that generally will not require a power cycle to clear.

 

[markav2]

I purchased my copy on April 18, 2010, and this is the first time out of many many uses that I have run into one of my drives that is so bad that SpinRite has run over 90 hours and has only completed a little over 27% of the 320 GB drive.

 

[peterblaise]

". . . Level 4 was changed to be Level 3 + an additional read. Level 5 is basically what Level 4 used to be . . . SpinRite 6.1's levels: . . . 4 Read, Write [ original ], Read 5 Read, Write inverted, Read, Write original, Read . . ."

THAT'S why LEVEL 4 cannot identify missing storage on a fake drive where LEVEL 5 can.

Thanks.

 

[peterblaise]

". . . I purchased my copy [ of SpinRite 6 ] on April 18, 2010, and this is the first time out of many many uses that I have run into one of my drives that is so bad that SpinRite [ 6.1 SRPR-RC3 ] has run over 90 hours and has only completed a little over 27% of the 320 GB drive . . ."

I use supplemental cooling fans to keep a drive at room temperature, especially with the added 'stress' of heat generation during SpinRite's constant drive analysis. I tend to run SpinRite 6.1 at DYNASTAT 1 limiting re-read attempts to 1 minute. There are millions of sectors, and 5 minutes for each one with a read error adds up. It's hard to compare the differences between the results of DYNASTAT 1 and DYNASTAT 5 because every sector and it's problems, if any, is unique. I also watch all the places @Steve Gibson posts in order to keep current, and I stop using prior pre-release versions of SpinRite 6.1 as soon as a new version is available, ( today, SRPR-RC5 ). I have found that no matter what I do, some drives need attention, some 'get better', some resist improvement, and some decay. SpinRite is essentially for data recovery and drive maintenance, and we decide what are our own acceptable thresholds for those.

Though your drive may be beyond usefulness for you, we all might learn something using the SpinRite 6.1 command line option of FORCEBIOS, which may 'encourage' the drive to address bad sectors without the drive going offline - try that and let us know.

Thanks.

 

[ShadowMeow]

Peter, have you tried FORCEBIOS yourself? Try that and let us know. Thanks.

 

[Mervyn Haynes]

and we decide what are our own acceptable thresholds for those.

You have changed your tune. Thought you wanted SR to to say "yay" or "nay" about a drive.

 

[peterblaise]

". . . Peter, have you tried FORCEBIOS yourself? Try that and let us know. Thanks . . ."

Oh yes, great point - never make a recommendation in speculation, only from experience.

GRC SpinRite 6.1 development resources are replete with my own FORCEBIOS explorations with @Steve Gibson, using that command line option on every computer in my suite as we test the software, that's why I suggested - from experience - that FORCEBIOS may get through shy sectors that otherwise throw a drive offline, where the IDE/ATA/AHCI drivers - and S.M.A.R.T. - get caught, there's an incredible history across the various GRC SpinRite development forums, newsgroups, and @Steve's GitLab on all of this.

There are 3 goals here at least during pre-release:

• test SpinRite 6.1 in development
• test data recovery
• test drive maintenance

FORCEBIOS is just one tool in SpinRite's toolbox, and it seems a worthy feature that may have the benefit of recovering and maintaining the shy sectors on the drive under test.

So why not?

I don't have possession of the drive, so I cannot try it on your drive, only you can.

So, let us know how FORCEBIOS works for you.

---

Then we can also discuss other command line options, like XFER n, where I successfully have gotten thousands of sectors deeper into shy areas using XFER 1, for example.

And changing the CMOS BIOS from AHCI to ATA or IDE/Legacy/Compatible/Native, and trying SpinRite 6.1 pre-release, plus trying any combination of command line options along with those CMOS BIOS changes - even moving to an older, slower SATA1 computer, and, of course, swapping SATA cables.

Thanks.

 

[peterblaise]

". . . You have changed your tune. Thought you wanted SR to to say "yay" or "nay" about a drive . . ."

I can't imagine SpinRite saying "yay" or "nay" to a drive, considering that SpinRite does not ask me what are my criteria for a drive being appropriate for or not.

There are two main features of SpinRite:

• data recovery
• drive maintenance

. . . plus inventory for the chips and connections, and Benchmark, I guess.

Even if SpinRite can't recover data or maintain a drive, that drive still might be appropriate for my needs:

• I may have a backup of the data anyway, so I don't need data recovery.
• Windows CHKDSK /R might confirm a drive is reliable for data storage and retrieval, and may mark as bad any clusters containing sectors that are too slow to respond, leaving the usable partition area happy and responsive.
• An inordinately slow drive may be excellent long-term storage of backup data - WORM write once, read many <-- SMR drives.

As folks who have reviewed my reports and LOGs know, I don't throw anything out, finding a use even for 'compromised' stuff somehow.

---

Perhaps you are thinking of my striving to get SpinRite 6.1 to recognize fake drives, being chagrined that LEVEL 2, 3, and 4 tests do not.

But LEVEL 5 does, so I'm satisfied that the fake-finding 'skill' is somewhere in SpinRite 6.1, if we dig for it.

Is that what you were thinking when you thought I wanted SpinRite to say "yay" or "nay" on a drive - fake drives?

Please find whatever reference made you think that and quote it to me.

Thanks.