The 1TB SSD raid for the /opt/rtcds file system is at 90% full. 513G of this is in the target_archive area. I'm currently creating compressed tar images of these directories (and then deleting the source). This will take several hours to complete. (these directories compress by a factor of roughly an order of magnitude).
Yesterday I compressed most of /opt/rtcds/lho/h1/target_archive. Because we are keeping about a month of hourly ZFS snapshots on h1fs0 (and its backup machine h1fs1) the recovered disk space is not immediately available. To offset the additional space needed for the compressed tar files, I manually deleted ten days of snapshots from 10th - 20th November. As the archived snapshots get rotated out, the disk space used will fall dramatically. At that point I'll compress the remaining targets and extend the lookback to a month again.