Stange effect of partition on drive performance

[paul1149]

I just had a strange experience I don't know how to interpret. I got a used laptop that was distressed and would not boot. I pulled the 500GB SATA II drive and ran some tests on it, and was dismayed to find transfer rates of under 20MB/s. I ran CHKDSK on it, which produced no errors and did not change the performance. The drive passed all SMART and self-tests.

So I put the drive back in its machine and began to try to repair the boot. That led to trying to do a factory recovery, but that also failed. Finally I decided to install Win7 afresh from a DVD, and as part of that I deleted the two existing installation partitions and let Windows create new ones. That went very smoothly, and after all the dust settled I ran some performance tests on the drive again. Now the drive was clocking somewhere in the 70MB/s range!

Great news, but I'm trying to figure out what changed. The machine had been dropped, so I feared HDD damage. There is one CRC SMART error, but aside from that everything is perfect. I don't know what reformatting the partitions has to do with drive performance.

 

[TVeblen]

Probably a bad sector that had been written to previously, then damaged. If the damaged sector contains data the OS needs then you have trouble. When you run chkdsk /r or format a drive the system will try and mark bad sectors as unusable and ignore (skip) them. So your new install will have all the data it needs available.

Probably a good idea to run the manufacturer's diagnostic - extended test - on the drive to be sure. I have seen drives pass the SMART checks but fail the extended tests.

 

[paul1149]

Thanks, but the performance I cited had nothing to do with the OS. It was raw testing of the drive itself. And the self-tests I mentioned included the extended test. That's exactly why this situation puzzles me.