Dropped PC - Bad HDD?

[kevin!]

Re: Dropped PC - Bad HDD?

Your drive not only has 87,157 power on hours but 2,348,146 power cycles. That's over 2 million power cycles! Don't you think that's a tad unhealty?

I would also recommend replacing the OP's drive. Modern Seagate drives (desktop drives at least) aren't very good anyway.

It was previously as a removable drive Conceptronic, and that fed PSU failure after 5 years of use, and every 3 minutes off and turned on the disc.
Now I have it mounted on my server without hdd case with PSU FSP, and without any corrupt sector. Good Seagate
Many people have had problems, and more with the 7200.12 series,
but I had one and never entered BSY.
and a case of a family that has a 500GB Maxtor has 22.000h, and no faults.

Personally I prefer Seagate over WD in recent years. 7200.11 and 7200.10 series were crap though. I have deployed many ST1000DM003 through clients, friends, personal computers, and work and have had good luck. The few WD10EZEX I have used have been good as well.

The older WD Blacks and Greens are absolute garbage. The latest ones are fine. It bothers me they have a half dozen series now though.

I have also been a lot of Seagate and is the brand that is best in HDD.
I've also had WDC, but his performance in multitasking was appalling.

All drives get read errors. They are going to increase over time. The SMART read error rate is basically useless information as the links that Per gives show. In addition, it doesn't tell us what sectors are associated with the errors, it could be the same sectors giving read errors over and over again. It does not indicate the drive will fail soon. Of course back up data important to you, which should be done anyway, continue to use the drive and keep an eye on it. I have seen LOTS of drives suffer a fall (in and out of computers) and keep working just fine for years. If you want to replace the drive anyway of course that is your choice.

I agree with you, many times you can die without warning the HDD smart.

Bye (sorry my bad english)

[ChaosLegionnaire]

Re: Dropped PC - Bad HDD?

the only way to know for sure is to run a benchmark NOT an error scan judging from the error. excessive rises in the seek and read error rate and having a smart failure in the seek error rate primarily mean mechanical failure and/or mechanical defects in the drive head and/or actuator seeking to the wrong position thus resulting in a seek error.

bad sectors typically mean defects in the disk surface or the magnetic bipoles of the hard disk platters. damage to the disk surface is unlikely when the drive is off since the heads are in the park position. its more likely the head or actuator got jarred by the shock of the drop instead and developed mechanical issues. this is considered a type of mechanical failure of the hard drive and the hard drive should be replaced.

this can be proven by running a benchmark of the drive like i said. the speed graph should spike up and down and the seek time benchmark result should be abnormally high. sometimes the drive may even freeze and cause the test to be aborted with a read error msg.

i have a hard drive with the exact same issues as shovenose. the hard drive is considered toast if it exhibits the spikes and high seek times regardless of whether there are bad sectors or not since the hard drive will freeze and lock up your programs/games while using/playing them. its unusable and unplayable like this.

i'd be really pissed at your friend if he didnt compensate u for at least part of the cost of a new hard drive. if u want really reliable, then hitachi pwns all from the backblaze hard drive study.

[mathog]

Re: Dropped PC - Bad HDD?

SMART won't show any bad sectors until it reads them.

This varies. Some drives will keep error counts when they write a sector, it fails, they swap it out, etc. until they find a good spare sector. Others don't. So two disks can read "no errors" even though one is down to its last spare block and the other hasn't touched a single one. Some WD's are like that. I have seen these on linux systems where they fail a SMART test, and I force it to swap it out by writing to that block with "dd", and when it is done all the relevant error counters are back to zero. Others keep accurate counts, so you can tell a good drive from a bad one.

Regarding the "rising error rate", I took that to mean that it was rising faster than before, that there had been some change. It is true that Seagate drives accumulate a huge number of certain types of errors normally.

[shovenose]

Re: Dropped PC - Bad HDD?

hm... well might not be getting a new drive dont really want to put my friend on the hook for something yes it was his fault but i am the idiot that loaded the computer on the edge of the cart.

[momaka]

Re: Dropped PC - Bad HDD?

st2000dm001

Errors in readings Seagate is a wrong value.
All seagates always worth reading errors are rising steadily, and this does not indicate a failure.

^ What Kevin said.

Raw Read Error Rate and Seek Error Rate can be ignored for Seagate hard drives. They will always show very high and ever-increasing values. In fact, I think I remember reading somewhere that these parameters were used to indicate something else for Seagate HDDs.

Either way, don't worry about those. Only look for increasing values in Reallocated Sector Count, Current Pending Sector Count, and sometimes Offline Scan Incorrect Sector Count.

That said, I would also like to point out that Seagate HDDs (especially older ones) often tend to develop a few bad sectors over time without affecting the reliability. I have quite a few Seagate HDDs like that - from Barracuda ATA-IV to 7200.10 series. I think I even posted the SMART logs for a few here (will look in my attachments if anyone is interested).

the only way to know for sure is to run a benchmark NOT an error scan judging from the error.

Yes, I find this to be true as well.
Have seen many HDDs (and from different brands) that were starting to go bad, and while some of them did not show anything with a surface read scan (like the free version of HDTune does), all of them acted weird with a benchmark test. Two of them, a Fujitsu MHW2120BH and a ST3500620AS, crashed. With the Fujitsu permanently (it had a head stuck on the platter that I got unstuck by opening the HDD, so I was expecting that HDD to fail).

[blasterboomer]

Re: Dropped PC - Bad HDD?

All seagates increase the raw read error rate, seek error rate, and hardware ecc recovered, its something that all seagates do for some reason so your drive is probably fine

[fzabkar]

Re: Dropped PC - Bad HDD?

SeaTools is useless as a SMART diagnostic. It simply reports pass or fail. A drive could have 2500 reallocated sectors, but SeaTools would still tell you that it is OK.

To test the surface for "slow" sectors I recommend HDDScan (Windows) or MHDD (DOS, better). I like CrystalDiskInfo for SMART data. HD Tune's author doesn't understand that raw SMART values are 48/56-bit numbers, not 32-bit. Also, many attributes, eg Seek Error Rate, are more meaningful when expressed in hexadecimal rather than decimal.

Seagate's raw Seek Error Rate, Raw Read Error Rate and Hardware ECC Recovered SMART attributes are sector counts, not error counts. The numbers for RRER and HER steadily increase and eventually roll over to 0 before increasing again.

Seagate's Seek Error Rate, Raw Read Error Rate, and Hardware ECC Recovered SMART attributes:
http://www.users.on.net/~fzabkar/HDD..._RRER_HEC.html

The author of the BackBlaze blog does not understand Seagate's attributes. He invites people to explain them to him, which I did, but he appears to be too proud to amend his article.

I'm still using a 40GB Seagate drive that has had 50 or so bad sectors for the past 2 years, without increase. Modern drives, OTOH, tend to fail quickly after developing bad sectors. AIUI, these bads are not usually due to media problems, but weak heads, probably as a result of too many "head slaps".