GFS / overlapping Snapshot policy

We currently have a lot of snapshots on our cluster and need to make it a bit more efficient for the longer term, ideally we need to keep snapshots for around 90 days.

Currently our general setup is as follows:

x2 a day, ~90 day/3month retention policy on most of our directories, which you can image equates to a lot of snapshots (180 per dir). ~10k at present on our cluster. I am aware there is a 20k limitation on snapshots on a single isilon cluster. 

We would like to ideally adjust our policies as follows:

x2 daily for 30 days

x1 week for 3 months

This would reduce our snapshots per dir from 180 to ~75. I know this might cause a slight increase in the total size of snapshots on disk, however I was more wondering if anyone knew of any overhead issues with overlapping snapshots?. I couldn't find anything too specific about it in the documentation. This is not overlapping in the sense of directories, eg. /ifs/data/dir1/folder1 and /ifs/data/dir1/folder1/folder2, but in time, so we have twice a day for 30 days of /ifs/data/dir/folder1 and then can also go back to 90 days via weekly snapshots of /ifs/data/dir/folder1.

our cluster is not under strain, so currently is coping well with the amount of snapshots we have, however we would like to get it closer to something more sustainable as the number of shares grows overtime. 

What do other users do with their snapshot policies? 

When deleting snapshots, this should always be done with the oldest first correct? 

what would happen if we move a dir around the cluster, e.g. mv /ifs/data/dir/folder1 /ifs/data/newdir/folder1 as i spotted some directories have been created in the wrong places. do snapshot policies need to be edited or deleted and started from scratch when you move a directory? 

Cheers,