Wikipedia:Reference desk/Archives/Computing/2018 August 18

= August 18 =

G-sensor on Western Digital mobile drives
Has anyone else noticed the G-sense error rate on WDC mobile drives being a little too sensitive? I've read that others have noticed this on what appears to be an otherwise perfectly fine laptop hard drive, most especially those from a brand new, bone stock device. Can anyone corroborate or clarify on whether the hard drive on my brand new HP laptop is going south prematurely or is there nothing I should worry about? Blake Gripling (talk) 06:14, 18 August 2018 (UTC)


 * What are the results when you run a WDC Data Lifeguard extended test? . --Guy Macon (talk) 06:32, 18 August 2018 (UTC)


 * Hmm, I'll see if that helps. Blake Gripling (talk) 06:39, 18 August 2018 (UTC)


 * Assuming that there are no problems detected, it may be helpful to understand what a G-sense error rate really is.


 * A G-Sense error usually means that the drive decided to abort and retry an operation because it detected an error at the same time that it detected shock or vibration. On some drives it means that the drive decided to abort and retry an operation because it detected shock or vibration without detecting an error. This is more desirable from a "just to be safe" standpoint.


 * There are four S.M.A.R.T. attributes related to this (not all disks implement all attributes):


 * Attribute ID 191 (decimal) / 0xBF (hexadecimal): G-Sense number of errors


 * Attribute ID 211 (decimal) 0xD3 (hexadecimal): Vibration encountered during write operation.


 * Attribute ID 212 (decimal) 0xD4 (hexadecimal): Shock encountered during write operation.


 * Attribute ID 221 (decimal) / 0xDD (hexadecimal): G-Sense frequency of errors


 * --Guy Macon (talk) 07:09, 18 August 2018 (UTC)


 * So that means it would record every single case of shock the sensor has detected, even minuscle ones? Is it something to worry about or should I concentrate more on racking up bad sectors? Blake Gripling (talk) 07:31, 18 August 2018 (UTC)


 * That's pretty much it. If the WDC Data Lifeguard extended test shows no problems and the S.M.A.R.T. isn't showing a lot of read errors or new bad sectors, I wouldn't worry about it. It isn't totally useless, though. I once had to troubleshoot an industrial robot that was throwing a G-Sense error on 20% of the reads because of vibration. Rubber shock mounts fixed the problem.
 * Your instincts are correct about worrying if there are to many bad sectors. There are two basic kinds of bad sectors; first there are minor defects from manufacturing. Those are normal -- nothing is perfect. Then there are new bad sectors that happen because the head hit the disk and kicked up a shower of particles, which then get between the head and the disk, causing another head crash and more particles. The drive does filter the air that circulates inside the drive, which catches a lot of the particles, but not all. --Guy Macon (talk) 10:35, 18 August 2018 (UTC)


 * I see a previous handling or assembling problem of the drive. The errors seem to describe a rough use of a mobile computer on the desk but no dropping from the to floor. It might also be caused by a wrong used rubber or spring buffer, enhancing shocks and vibration. Usually the drive bay in a case should be stable to reduce vibration and the case should eat the shocks to protect the drive bay with all drives inside. -- Hans Haase (有问题吗) 12:50, 25 August 2018 (UTC)