I'd just like to ask a question regarding on which Surface Test type should I use for scanning my HDD for possible CRC errors. I noticed that my hard drive started to get "Ultra DMA CRC Error" on S.M.A.R.T ever since I upgraded my laptop's RAM with a defective one.
The "Ultra DMA CRC Error count" S.M.A.R.T. attributes reads "3" and I noticed it all happened after my laptop refused to boot Windows 7 for three times.
Right now, I have the faulty RAM replaced and all is well now, no more weird stuff happening on boot and the RAM and my HDD passed their respective tests.
Still, I'm feeling a bit paranoid despite finding out that the defective RAM caused my drive to get CRC errors...and I still want to know if there's really something wrong with my HDD.
I'd also like to ask if it's really possible for a bad stick of RAM to cause CRC errors on a hard drive and how does it happen (for example, do data bounce back and forth between the RAM and HDD during read and write operations?) .
And lastly, which Surface Test type do you recommend for scanning my drive for CRC errors. I have created a backup of all my files and I'm waiting for your excellent recommendation before doing the test.
Thanks and more power to you and your almighty software that saved a lot of drives in the world!
Which Surface Test type to use for scanning CRC errors?
-
hdsentinel
- Site Admin
- Posts: 3128
- Joined: 2008.07.27. 17:00
- Location: Hungary
- Contact:
Re: Which Surface Test type to use for scanning CRC errors?
I completely agree: personally after similar incidents, I'd perform many different tests to verify the hard disk status and the operation environment in general, to confirm that things are fixed - and there will be no more troubles in the future.
Generally, the Disk menu -> Surface test -> Read test (the complete read test) is a good way to verify if the complete surface is accessible. During the read test, most of the possible CRC errors can be detected.
But if really want to be sure, I'd recommend the Disk menu -> Surface test -> Read-write-read (Refresh data area) test in this situaiton. This would read the contents of each seectors, overwrite with initialisation pattern, write back the original data and read back + compare with the information previously written.
So this would surely point on
- any kind of CRC errors (even if they happen during reading or writing)
- any kind of data corruption, for example the data somehow gets corrupted / modified (for example because of the faulty RAM).
(note that for this type of test you'd need to use the hard disk as secondary storage: there should no software / OS running or using the hard disk when you prefer to start the test. But as you wrote, you have a complete backup, so I assume it is possible).
> I'd also like to ask if it's really possible for a bad stick of RAM to cause CRC errors on a hard drive
Yes, it can be the source of such problem as well - as then the operating environment is not stable.
In some rare cases even overclocking can cause CRC errors as it may cause synchronization issues.
(for more information, you may check www.hdsentinel.com -> Support -> Knowledge Base -> Hard Disk Cases -> Communication Errors (or directly at http://www.hdsentinel.com/hard_disk_cas ... _error.php )
> and how does it happen (for example, do data bounce back and forth between the RAM and HDD during read and write operations?) .
Yes: and this results degraded performance and in some cases even system crashes / BSODs / sudden restarts (but these latters more likely happen with older IDE/PATA hard disk drives).
Generally, the Disk menu -> Surface test -> Read test (the complete read test) is a good way to verify if the complete surface is accessible. During the read test, most of the possible CRC errors can be detected.
But if really want to be sure, I'd recommend the Disk menu -> Surface test -> Read-write-read (Refresh data area) test in this situaiton. This would read the contents of each seectors, overwrite with initialisation pattern, write back the original data and read back + compare with the information previously written.
So this would surely point on
- any kind of CRC errors (even if they happen during reading or writing)
- any kind of data corruption, for example the data somehow gets corrupted / modified (for example because of the faulty RAM).
(note that for this type of test you'd need to use the hard disk as secondary storage: there should no software / OS running or using the hard disk when you prefer to start the test. But as you wrote, you have a complete backup, so I assume it is possible).
> I'd also like to ask if it's really possible for a bad stick of RAM to cause CRC errors on a hard drive
Yes, it can be the source of such problem as well - as then the operating environment is not stable.
In some rare cases even overclocking can cause CRC errors as it may cause synchronization issues.
(for more information, you may check www.hdsentinel.com -> Support -> Knowledge Base -> Hard Disk Cases -> Communication Errors (or directly at http://www.hdsentinel.com/hard_disk_cas ... _error.php )
> and how does it happen (for example, do data bounce back and forth between the RAM and HDD during read and write operations?) .
Yes: and this results degraded performance and in some cases even system crashes / BSODs / sudden restarts (but these latters more likely happen with older IDE/PATA hard disk drives).
Re: Which Surface Test type to use for scanning CRC errors?
As expected from you, the creator of this wonderful software, your response is fast and very comprehensive! No wonder many people call your software as the best Hard Disk utility in the world!
I've done a read test last night and fortunately, the test did not return any errors and the SMART values stayed the same. Here's hoping that the test I'll be doing tonight will not return any errors too.
I've done a read test last night and fortunately, the test did not return any errors and the SMART values stayed the same. Here's hoping that the test I'll be doing tonight will not return any errors too.