Is the volume make-vsroot command disruptive to SVM volume junction paths?
Applies to
Answer
Yes.
- The
volume make-vsroot
command is used to promote a non-root volume of the respective SVM to be the SVM's new root volume. - This new SVM root volume must be a FlexVol volume with a size of at least 1 GB.
- Whenever a volume is promoted to be the new root volume of the target SVM, all junction paths to all associated volumes under the SVM will be disconnected, which will induce a service outage for the affected volumes.
Additional Information
- Create a new target root volume for the respective SVM, i.e.
::> volume create -volume <new_svm_root> -aggregate <aggr name> -size 1GB
- After the volume is created, enter diagnostic mode with
::> set diag
- The while in diag mode, execute the command to promote the new SVM root volume with
::*> volume make-vsroot -vserver <SVM name> -volume <new_svm_root>
- Check the new SVM root volume with this command to insure that it has been successfully promoted:
::> volume show -volume new_svm2_root -fields vsroot
vserver volume vsroot
------- ------------- ------
svm2 new_svm2_root true
Other Notes: -
If unmount fails, you may have to use a "volume offline" and subsequent "volume online" for each volume that fails.
-
Make sure to match the export policy of the prior SVM root with the "-policy" on the volume create to be sure there are traverse RO permissions. (the default policy has no rules unless it was modified)
-
Check and create any directory paths if the volume mount paths are not "/{volname}". Some volumes might have been mounted using "/vol/{volname}"
-
Be sure to create new LS Mirrors for the new SVM root volumes.