The error above has been seen when a disk that was previously used on a different system managed by Solaris Volume Manager (SVM) and possibly Sun Cluster. In this scenario, the following occured:
- Volume Manager detected the disk as "auto:SVM" in the output of vxdisk list (vxdisk -eo alldgs list).
- since the disk was used by SVM on a different system, the customer worked with Oracle at the OS layer to remove the SVM artifacts from the disk so that the disk would be detected as "auto:none" (after removing artifacts, run command 'vxdctl enable')
Please engage Oracle to remove SVM artifacts from the disk and/or OS configuration as appropriate so the command 'vxdisk list' will report a TYPE of "auto:none". If the disk was also use in a Sun Cluster environment, there may be stale keys left on the disk (SCSI3 persistant reservation). Please check the disk to see if these exist and remove as needed.
Applies To
While this issue was dicovered on Solaris 10 running Volume Manager 5.1SP1RP2, it could likely occur in any Solaris environment running any supported version of Volume Manager (Storage Foundation).