Experience with CHKDSK /R?

[codegreen]

I agree with atlanticjim -- skip the third option. To fall back to the ever-popular car mechanic analogy, if you need a new muffler, a professional mechanic isn't going to offer a third option involving duct tape and super glue to "fix" your current muffler. You either get the muffler replaced, or you pay the diagnostic fee and take your car. Likewise, you shouldn't offer a "fixed" hard drive that could crash at any time. You don't want your name attached to that kind of work.

[PCX]

The third option should not even be considered an option. Its completely unprofessional and it will come back to haunt you eventually. Do not do anything that you cannot or do not feel comfortable warrantying, that is of course with the exception of software repairs like virus removals.

[compnet]

Or like if you take a car to a mechanic with a nail in the tire. He's not just gonna patch and glu- oh wait... He'll probably use his professional opinion to decide if a patch will hold or if the tire needs to be replaced. Just because you always replace the tire doesn't mean he is wrong. (and technically mufflers can be patched safely, along with tons of other car issues)


Chkdsk when used properly is not dangerous, that's just dumb. Were talking a couple bad sectors. It most likely had more than that remapped from the factory.

This thread is getting into the whole bad sector debate again.. if you have thousands of bad sectors and is failing real diagnostics, and your first step is chkdsk /r ya you could lose some data. If there is one bad sector, it won't boot because that's part of a system file, and you get them back up and running in an hour, with no proven greater chance of failure, then you are saving them time and money with no down side. Of course recommend backing up and alert the customer that the drive did fail a minor diagnostic, giving them the opportunity to replace.

Half the arguments in this thread could be used to argue a nuke and pave should be performed for any software issue, virus, etc.

[iisjman07]

Originally Posted by compnet

Or like if you take a car to a mechanic with a nail in the tire. He's not just gonna patch and glu- oh wait... He'll probably use his professional opinion to decide if a patch will hold or if the tire needs to be replaced. Just because you always replace the tire doesn't mean he is wrong. (and technically mufflers can be patched safely, along with tons of other car issues)


Chkdsk when used properly is not dangerous, that's just dumb. Were talking a couple bad sectors. It most likely had more than that remapped from the factory.

This thread is getting into the whole bad sector debate again.. if you have thousands of bad sectors and is failing real diagnostics, and your first step is chkdsk /r ya you could lose some data. If there is one bad sector, it won't boot because that's part of a system file, and you get them back up and running in an hour, with no proven greater chance of failure, then you are saving them time and money with no down side. Of course recommend backing up and alert the customer that the drive did fail a minor diagnostic, giving them the opportunity to replace.

Half the arguments in this thread could be used to argue a nuke and pave should be performed for any software issue, virus, etc.

I disagree.

With car, if your tyre has a nail in it and you need a new tyre, once it's fitted and paid for, most people probably wouldn't notice the difference to when the previous tyre was functioning. Going on the basis that nobody backs up their years of digital memories and computers are cheap nowadays, your number one priority and the most valuable part of a computer is the data on it. The vast majority of people couldn't care less about their computer, just their financial & business, digital photos, music, videos, and so much more. By not replacing a failing hard drive you could potentially jeopardize the client's valuable information in the future. It's the professional thing to do.

Half the arguments in this thread could be used to argue a nuke and pave should be performed for any software issue, virus, etc.

All things considered, you can never be truly sure an infection has been removed with nuking the drive (especially with recent secretive rootkits, keyloggers, etc). If a sysadmin had a server that was infected, it's best practise to rebuild it from scratch. To be sure an infection is gone, can't return, and the client is protected, the only method is to nuke, so maybe it is the most professional thing to do.

[gikstar]

Originally Posted by techyguy717

I ran chkdsk /r on a Windows drive that detected a couple bad sectors. (Windows was freezing)

It notified that it found and fixed all errors successfully.

What are people's success rates with this when ALL bad sectors ARE FIXED?

Because my understanding is that bad sectors could spread causing even more drive failure. Alternatively could a bad shutdown or power outage cause bad sectors, therefor causing the drive to still be good in all other areas? In which case a chkdsk /r would be a permanent fix.

Would it be appropriate to offer to the customer this solution, but notify that the only way to make sure the problem is fixed is to buy a new hard drive.

1) 1 Hour Charge = possible fix
2) 3 Hour Charge + $100 Hard Drive guaranteed fix and backup.

And what would you charge if you were successful in using Acronis to clone the old drive to the new drive and verify if the system is running properly?
Also with option 2) what would be the total 'fee' charged?

[PCX]

Originally Posted by compnet

Or like if you take a car to a mechanic with a nail in the tire. He's not just gonna patch and glu- oh wait... He'll probably use his professional opinion to decide if a patch will hold or if the tire needs to be replaced. Just because you always replace the tire doesn't mean he is wrong. (and technically mufflers can be patched safely, along with tons of other car issues)


Chkdsk when used properly is not dangerous, that's just dumb. Were talking a couple bad sectors. It most likely had more than that remapped from the factory.

This thread is getting into the whole bad sector debate again.. if you have thousands of bad sectors and is failing real diagnostics, and your first step is chkdsk /r ya you could lose some data. If there is one bad sector, it won't boot because that's part of a system file, and you get them back up and running in an hour, with no proven greater chance of failure, then you are saving them time and money with no down side. Of course recommend backing up and alert the customer that the drive did fail a minor diagnostic, giving them the opportunity to replace.

Half the arguments in this thread could be used to argue a nuke and pave should be performed for any software issue, virus, etc.

You have to be your own judge on what is too many bad sectors and what is not. I will tell you that you will not always be right. Some times 1 bad sector is enough before the hard drive takes a nose dive. Is it likely, no, but possible. In my experience, around 10 is usually fine. Chkdsk /r is probably an appropriate move at this point, but be sure to backup their data just in case. It is also in my experience that once you start getting around 20 bad sectors, due to the huge increase in read failures, you need to suggest a hard drive replacement. If you are trying to run chkdsk /r with a ton of bad sectors or read errors, then that is just stupid . . . unless you have cloned the hard drive and you are running chkdsk /r on the new drive. As others have already stated, the most important thing is their data.

[Dean]

I always do a hardware check first with crystal disk info etc. if the drive is ok then I will run chkdsk /r. This works great with non boot issues

[iisjman07]

Originally Posted by Dean

I always do a hardware check first with crystal disk info etc. if the drive is ok then I will run chkdsk /r. This works great with non boot issues

You shouldn't have said that when PCX is around...

[PCX]

Originally Posted by Dean

I always do a hardware check first with crystal disk info etc. if the drive is ok then I will run chkdsk /r. This works great with non boot issues

Chrystaldiskinfo is not a diagnostics tool that will tell you if a hard drive is failing. It is a monitoring tool that will "monitor" a hard drive over time and hopefully warn you of possible hard drive failure before it does fail. It only reads SMART data, it does not actually test the hard drive. Unless the hard drive has failed in the past, the SMART data will probably not show that the hard drive is currently failing.

The point is, with chrystaldiskinfo, unless the drive is severely failing or has failed in the past, you will not know what the drives current state is and you are taking a huge risk with your customers data. I cannot count the number of times a drive has checked out on the SMART data and then failed the SMART tests.

[PCX]

Originally Posted by iisjman07

You shouldn't have said that when PCX is around...

Man you are quick on the trigger! lol.