Storage Solutions How many bad sectors are okay?

[Kneo]

I have a 500Gb Seagate HDD (RMA'd once - received refurbished HDD) which is in warranty until 2013.

This HDD has started developed bad sectors and shows up 10 'red' blocks in HD Tune. I have quite a bit of data on it which I don't want to lose and I don't have any spare HDD to back that data up. I'll be able to live with the 10 blocks - as I don't want to go through the process of RMA (lack of time to stand in the queues).

Is there anyway I can ensure that the bad sectors are cordoned off and not used anymore?

 

[..:: Free Radical ::..]

If your drive is in warranty, it is utterly foolish to not RMA if it has developed bad sectors.

For old hard disks which are out of warranty and have bad sectors, use MHDD.

Its an advanced low level tool so be very careful with it and read up on how to use it.

You'll have to boot it off a pen drive and then you can do a zero fill and perform reallocations of the bad sectors and label sectors with high latencies as bad so that your drive never writes on those area.

But before you do anything, backup your data on a good disk / optical media.

I'd recommend buying a spare hard disk. The are quite cheap nowadays and its always better to have more storage.

 

[Kneo]

Would a HDD regenerator be useful?

 

[..:: Free Radical ::..]

Nope. These HDD regenerators simply relabel bad sectors as good after doing a certain number of rewrites and reading the data written overlooking latencies. That can only cause further problems.

 

[blr_p]

What about good old chkdsk with the repair option to move data from bad to good sectors and label the bad ones so they're not used again. It does not do the latency thing you mentioned tho.

Re: the topic's title, i vaguely recall that windows sets some limit for bad sectors which once passed then refuses to read the partition. That does not necessarily mean you lost it just that windows wont read it anymore, a bootup linux will.

 

[..:: Free Radical ::..]

What about good old chkdsk with the repair option to move data from bad to good sectors and label the bad ones so they're not used again. It does not do the latency thing you mentioned tho.

Re: the topic's title, i vaguely recall that windows sets some limit for bad sectors which once passed then refuses to read the partition. That does not necessarily mean you lost it just that windows wont read it anymore, a bootup linux will.

Three words:

"file system errors"

The /r switch mainly recovers readable information. It doesn't relabel bad sectors as good.

 

[blr_p]

The /r switch mainly recovers readable information. It doesn't relabel bad sectors as good.

Did you mean to say it does not label bad sectors as bad ?

..in which case those bad sectors would be reused again.

Hmm..in another thead someone recommended a format to get those sectors labelled properly. Which is much more work tho.

There is one downside i can see with using chkdsk is it will dump whatever is readable in a folder called FOUND.001 so if file integrity is the aim then the recovered files using this method are as good as useless if they are binary.

Does your utility (MHDD) retain file integrity assuming it was successful in reading out a file that spanned a bad sector ?

 

[..:: Free Radical ::..]

nope. I meant that if a sector is flagged as bad, at max chkdisk can perform a reallocation from spare sectors on your hard disk (added for fault tolerance and can number from a few hundred to a few thousand depending on the quality of the hard disk)

It does not mark the bad sector as good so its safe to use unlike the so called "hard disk regenerator" program the op mentions.

The FOUND* directories contains the raw dump of the erroneous files (saved as *.chk) whenever chkdisk encounters an error reading a sector which that particular file points to, and yes they don't have any any practical use.

MHDD does not distinguish between files (infact it overwrites them .... zerofill. duh!) You must take a backup of your data before performing it as it totally wipes your disk clean. It just reads and verifies individual sectors and shows the time taken (latency) to read a block (writes a 0 on a sector, reads whether its written correctly and if so, how much time it took between writing and verifying), which if not within the permissible limit (which can be specified by the user) can be flagged as bad. Basically, sectors which have high latencies are prone to become bad sectors so you can take pre-emptive action and label them as bad to prevent data corruption i.e. to cordon off bad sectors or those which are impending failure. Not only that, you can actually specify a range of sectors to not write to, eg. if a particular part of the hard disk has a lot of bad sectors.

Useful if you are working on sensitive data and encounter silent corruptions or want to salvage an old hard disk. I had kept my old 40 gigger pata on life support using mhdd until I retired it.

Phase 4: Checking sectors

If the /R switch is in effect, CHKDSK runs a fourth pass to look for bad sectors in the volume's free space. CHKDSK attempts to read every sector on the volume to confirm that the sector is usable. Even without the /R switch, CHKDSK always reads sectors that are associated with metadata. Sectors that are associated with user data are read during earlier phases of CHKDSK if the /R switch is specified.

When CHKDSK finds an unreadable sector, NTFS adds the cluster that contains that sector to its list of bad clusters. If the bad cluster is in use, CHKDSK allocates a new cluster to do the job of the bad cluster. If you are using a fault-tolerant disk, NTFS recovers the bad cluster's data and writes the data to the newly allocated cluster. Otherwise, the new cluster is filled with a pattern of 0xFF bytes.

If NTFS encounters unreadable sectors during the course of normal operation, NTFS remaps the sectors in the same way that it does when CHKDSK runs. Therefore, using the /R switch is usually not essential. However, using the /R switch is a convenient way to scan the entire volume if you suspect that a disk might have bad sectors.

 

[blr_p]

MHDD does not distinguish between files (infact it overwrites them .... zerofill. duh!) You must take a backup of your data before performing it as it totally wipes your disk clean. It just reads and verifies individual sectors and shows the time taken (latency) to read a block (writes a 0 on a sector, reads whether its written correctly and if so, how much time it took between writing and verifying), which if not within the permissible limit (which can be specified by the user) can be flagged as bad. Basically, sectors which have high latencies are prone to become bad sectors so you can take pre-emptive action and label them as bad to prevent data corruption i.e. to cordon off bad sectors or those which are impending failure. Not only that, you can actually specify a range of sectors to not write to, eg. if a particular part of the hard disk has a lot of bad sectors.

Useful if you are working on sensitive data and encounter silent corruptions or want to salvage an old hard disk. I had kept my old 40 gigger pata on life support using mhdd until I retired it.

Been playing with MHDD, its an interesting app.

Going through the docs the creator, recommends
- a zerofill (or ERASE)
- followed by a Scan + Erase Delays
- and finally a Scan + Remap

I'm not sure as to the point of the last step ?

I did the erase delays step and then ran the final and could see the brown & reds showing up again. So what am I remapping in this case ?

Also the prog uses a default of 350ms to keep out the red blocks with the longest latencies, is this ok, or do you suggest a shorter latency and if so how much ?

 

[..:: Free Radical ::..]

omg. tell me that you did that on a bad hard disk only!

Do not use SCAN + REMAP if checkdisk doesn't find any bad sectors.

<500 ms delay is good enough. But it really depends on your drive. The delays won't be removed unless you run the scan +remap after scan +erase

The remap switch reallocates bad sectors if there are any spare left and a remap event shows as a blue block.

HDD FAQ : Sticky/important topics. Beginners please look here first!

 

[6pack]

The remap switch reallocates bad sectors if there are any spare left and a remap event shows as a blue block.

doesn't the s.m.a.r.t on a hdd do that automatically? why use third party software for that? my now dead hdd had spare sectors remapped by the s.m.a.r.t automatically until no spare sectors were left and the hdd failed the s.m.a.r.t test.

imo, aren't all the hdd maintenance operations (except defragementation) taken care of by the s.m.a.r.t on hdd?

 

[Tech_enthu]

Does the mhdd also clear other faults of the hdd such as, spin retry count error, I have a brand new 1 Tb hdd and it shows spin retry count error, the windows chkdsk does not find anything wrong on the drive though and the seagate rma also did not find anything wrong with the hdd. so they wont give a new drive to me as seatools also does not give any errors...

 

[..:: Free Radical ::..]

doesn't the s.m.a.r.t on a hdd do that automatically? why use third party software for that? my now dead hdd had spare sectors remapped by the s.m.a.r.t automatically until no spare sectors were left and the hdd failed the s.m.a.r.t test.
imo, aren't all the hdd maintenance operations (except defragementation) taken care of by the s.m.a.r.t on hdd?

:S

S.M.A.R.T. = Self-Monitoring, Analysis, and Reporting Technology


Short answer: No.

It must be checkdisk which must have been run after a reported failure.
The op wanted to know a specific app which could be used to cordon off his bad sectors.

I won't recommend it to the average user, especially if you don't feel the need to make yourself familiar with such stuff.

 

[6pack]

^ no, checkdisk was not run on the drive any time.

This is copied from the wd diagnostic utility

Self-Monitoring, Analysis, And Reporting Technology

S.M.A.R.T. is an attempt to predict imminent drive failure by monitoring selected drive performance and calibration characteristics. Each monitored characteristic (Attribute) is used to calculate an Attribute Value. Attribute Values range from 1 to 253 with 1 being worst case, 253 being best case, and 100 being nominal.

For WD drives, the normalized attribute value will range from 1 to 100 or 1 to 200, depending on the attribute. Error rate, total spin ups, spin up retries and calibration retries attributes are implemented as during normal operation (“on-line”); whereas, Multi-zone Error Rate test is performed “off-line”. This test performs a read-verify-all operation and calculates an error rate. These Attributes are updated and stored on the drive in the reserved area of the disk. Also stored on the drive is a set of Attribute Thresholds that correspond to the calculated Attribute Values. An impending degrading or faulty condition is indicated when the calculated Attribute Value becomes less than or equal to its corresponding Attribute Threshold value.

The drive periodically saves all data associated with the S.M.A.R.T. feature. The data is written to the reserved area of the disk after certain time of inactivity by the host. The save of the data is performed in the background and will not cause a decrease in performance.

 

[..:: Free Radical ::..]

^

 

[blr_p]

omg. tell me that you did that on a bad hard disk only!

Do not use SCAN + REMAP if checkdisk doesn't find any bad sectors.

I ran HD Tune's error scan on an old 30GB drive which had data. Saw some damaged (red) blocks. So i imaged the data then formatted the drive in explorer. Ran HD Tune again and no red sectors showed up. So plain windows format does mark bad sectors and removes them from the lot. A chkdsk showed no bad sectors.

I then played with MHDD, did the erase and then the scan+erase delays.

Saw it writing in areas and when it completed then tried to do the scan+remap which I stopped a few minutes into the remap.

So if I understood you right then all thats needed with this drive is just the scan+erase delays with MHDD default setting for 350ms and thats it. Any potentially weak sectors will be written over so they cannot be reused.

Its interesting that even tho HD Tune showed all the drive as green, MHDD has written over 1-2% of blocks.

<500 ms delay is good enough. But it really depends on your drive. The delays won't be removed unless you run the scan +remap after scan +erase

The raw value for reallocated sectors count on this drive is 2 so thats fairly low.

Are you saying i need to do a scan+remap after the scan + erase now ?

The remap switch reallocates bad sectors if there are any spare left and a remap event shows as a blue block.

He says remap is safe so long as there are ample sectors remaining to reallocate which in this case there are.

 

[..:: Free Radical ::..]

you need to run scan +remap after scan +erase delays if you want to reallocate the sectors with delays.

Else, scan +remap after a zerofill to reallocate only the bad blocks

The scan option alone only reads and verifies the sectors.

The remap option reallocates sectors. But remember, you are consuming a finite number of spare sectors while performing reallocations so you won't lose any drive space as long as there are spare sectors, however your drive will start losing capacity after they are consumed.

If HDtune no longer shows any red blocks, you are good as reallocations have removed the bad sectors

 

[blr_p]

h

you need to run scan +remap after scan +erase delays if you want to reallocate the sectors with delays.

This is the bit I don't get

It's not marking the slow sectors with anything is it. The erase delays scan fills the slow sectors so they cannot be used without marking them as bad.

So what point is there to follow up with a remap in this case ?

The drive has just been erased (zero filled), and the slow sectors have also been filled

The other point with remap of bad sectors is also confusing. If a sector is bad, ithe dirve on its own marks it as such and nothing more gets done with it. What good will remapping in this case do then ?

Given the sequence matters, ie erase delays then remap. Tho some ppl on his forum tend to mix it up his FAQ clearly states remap follows after erase delays.

Then there is the makebad program which marks any of the slow sectors as bad so they're never used. I did not see much point with this as a filled sector is as good as useless. It's a better method to marking as bad and increasing the bad sector count.

What I like with MHDD so far is that its a quick way to check a drive with a scan and see its general health, that it talks directly to the HD without any intervening windows layer is a plus. A HDD program that talks to the drive directly via the BIOS will be more consistant with results than one doing it on top of windows

I guess the remap could help with restoring data from a bad sector by repeated retries. Maybe you get lucky or not.

What other unique uses does MHDD have for you ?

I'm currently using it on my worst 120GB drive, reallocated sectors through the roof, HD Tune marks this drive as failed (!), Its got tons of bad sectors (UNC + one other flag). But a surface scan shows them to be in a cluster so i'd made two partitions earlier and left the intervening 9GB which has all the bad sectors out. chkdsk on the two created partions yields no bad sectors.

So i'm giving this drive the full MHDD erase, scan & remap treatment to see what difference it makes with future scans. Other than the bad cluster in the middle the rest is not bad, very few greens even.

 

[..:: Free Radical ::..]

Then there is the makebad program which marks any of the slow sectors as bad so they're never used. I did not see much point with this as a filled sector is as good as useless. It's a better method to marking as bad and increasing the bad sector count.

this is exactly what we are doing with mhdd too.

I have even used MHDD when I could not detect the drive in my BIOS (you'll have to specify the location of the drive using shift+f3 (it brought it back to life after doing some remaps!)

Another time, when a power outage forced a part of my hard disk capacity to mysteriously vanish, a zerofill helped me get it back.

Since its been years since I last used MHDD, I'd recommend you to make your own inferences, but its nevertheless an amazing software and something I always keep on a bootable pen drive.

 

[blr_p]

this is exactly what we are doing with mhdd too.

You sure about this ?

As in it does it automatically, thought you had to specify a sector and run the makebad command explicitly on it. His forum isn't the greatest of places to get any info, lots of snarky comments from suposedly knowledgeable ppl.

Anyway, once the scans are done i'll take a look at the logs and see what the differences are.

I have even used MHDD when I could not detect the drive in my BIOS (you'll have to specify the location of the drive using shift+f3 (it brought it back to life after doing some remaps!)

Good to know, drive vanishing in the BIOS is really scary :O

Thing is if that happened then the reallocate sector count must have been very high. You then did a reallocate again with remap and yet it appeared. And you do say BIOS so its not like some bootable sectors got corrupted or anything.

Another time, when a power outage forced a part of my hard disk capacity to mysteriously vanish, a zerofill helped me get it back.

I've read posts where one person said that the slow sector count dropped signifncantly after a zerofill, it does have a tendency to rejuvenate things.

but its nevertheless an amazing software and something I always keep on a bootable pen drive.

Thx for recommending it