I want to find out when the cut-off point occur when using flexclone to migrate data from one vol to another.
Say I have a cifs volume cifsvol1, it contains 6 folders. I want to move two folders folder1 and folder2 out from this volume to another volume cifsvol2. All these folders are continuous accessed by different application.
My plan is
1) create a flexclone cifsvol2 from source volume cifsvol1,
2) remove folders folder3-6 from cifsvol2 so it only contains folder1 and folder2,
3) create new cifs shares for folder1 and 2 from cifsvol2,
4) split off clone vol
5) stop applications that are accessing folder1 and folder2, and remount them on the new cifs shares from cifsvol2
6) on the original volume cifsvol1, remove the cifs share for folder1-2, then delete them from volume cifsvol1
my question is, since flexclone creation is based on a snapshot, there most likely will be new updates from application write to folder1 and folder2 after clone volume is created, will these new writes be lost when performing step 5 above? how can I keep apps data intact? should I stop application before creating clone vol?
Yes, you have to stop applications before creating clone. Actually, I'd remove shares before creating clones as well to make sure noone can access these folders. After clone is created, recreate shares again pointing to new volume.
Thanks for the clarification, yes, we cannot afford to offline the entire cifs volume, so I think I will go for abrzenkov's suggestion.
Because this process requires downtime, that is from the time shutting down application, split off the clone volume, then remounting to the new cifs shares, I know creating flexclone should only take seconds, but it may take a bit of time to split off, I am wondering if there is way to work out how long it will take for splitting a 40G clone volume?