Sun Storage 3310 Array
 

GCS>Sun Microsystems>Storage - Disk>Modular Disk - 3xxx Arrays
 

A second SE3310 is configured the same way and mirrors the data in VxVM.

For some time now, the array has been reporting the same 4 media errors on ID1 every 4 days. The errors are always unrecoverable by the array of course because here is no data redundancy.The errors are being generated by media scan on the array The errors however are *not* passed up to the host, and we have no way to correct them. The errors have no affect on the filesystem.

When the the area on the drive is written to, the drive will automatically re-vector those LBAs and the errors will not occur.

The errors can just be ignored until the filesystem attempts to write them, and at which time they will automatically be revectored.

It can be very annoying to keep seeing the same event messages occur every few days, so that may not be the best option.

Assuming that the partition containing those blocks is in the filesystem, repairing it will be a bit difficult. If you want to attempt it, you will have to remove it from the metadevice, write to that area (with dd for example), then add it back to the metadevice and resync.

If the bad blocks are in a partition that is not in use, you can just overwrite them on the raw device.

That may be more trouble than the customer is willing to do. If so, you may have to replace the drive. It would be safe to use that drive as a spare, so if your configuration contains a spare, that might be a good option. If that new spare is used, the rebuild process will overwrite those blocks and correct the fault. You will not be able to use clone and replace, because the RAID controller will label those blocks as bad on the replacement drive, it will have to be replaced by the volume manager.

This solution has no attachment