Hello,

After a power fall (for some reason the APC UPS didn't helped) the sever (4.7) went down. Trying to bring it up failed. The error message goes like that:

ata4.01: failed command: READ DMA

.....

....

ata4.01: status: { DRDY ERR }

ata4.01: error: { UNC }

 

and so on.

Please see attached screen picture.

Do you think it's MotherBoard problem? I also  have a separate SATA PCi card. uninstalling it didn't help. Tried Memtest (didn't though that was the problem, nevertheless tried it.) Memory is OK.

Help will be much appreciated.

Thanks

Haim

Zip up and post a full syslog. It looks to me like you have a drive failing.

It is not a DMA error, that just means that reads are using the DMA channels, as they should, completely normal.  And it has nothing to do with the motherboard or memory or PCI.

 

The important flags are the 'media error' and 'UNC' (UNCorrectable sector info), which means you have bad sectors which need fixing.  Add a SMART report for the drive associated with ata4.01 to the syslog you post.  If you aren't sure which one that is, try collecting SMART reports for all of the drives, and include in the zip.

 

This may be the best kind of unreadable sectors, if they were caused by a power spike (at the beginning and/or end of the power outage) while the drive was being written.  That tends to scramble data, making the sector unreadable, but usually the platter media is not damaged, so there isn't anything wrong with the drive.  Forcing a drive rebuild causes the data to be rewritten in place and clears the errors.  Hopefully that is all you will have to do.

Hello,

Thanks jonathanm and RobJ, I was off line for a few days and couldn't check the forum.

Well, why a drive failure like you suggested prevents the system from booting?

How do I perform SMART tests without being able to boot up the system?

thanks

haim

Apparently all I have to do is to wait for several moments until the system will boot. Now I can handle the defect.

Thanks