while the snapshot size is kept at the level specified by the settings, the workspace size keeps growing. I used to go around and clean them up once a week but since I've been monitoring them, it's been more and more often that I run out of space.
Initially I noticed that some kind of cleanup kicked in at around 40 GB of disk space used, but right now the two nodes that have nearly unlimited disk space (and I haven't cleaned for a while) are at 80 GB and 64 GB, respectively. Is there any way we can restrict the disk usage of the workspace folder?
I recently found a flaw in the workspace cleanup code on the grid node. The node would still clean up workspaces, but it wouldn't do this properly on client disconnection. If the node was recycled, it would kick in with a cleanup, but this would require a manual action.
A fix for this has been released in NCrunch V3. I recommend giving the new build a go to see if this solves the problem for you.