Replies: 3 comments 4 replies
-
We have multiple harvesters with many drives per machine. They fill up at slightly different speeds but all use the same docker container to plot. In the future, we will have a problem of moving plots around for efficiency. We don’t want to, for example, launch a container to plot to drive 15 if that drive is already full in some of the machines. We don’t want to use raid zero for final directories either because each machine has so many drives. Is Chia or MadMax planning to support multiple final directories or is there a good tool? Something that checks and creates a new plot to whichever final directory still has space. Any help is appreciated. Too many machines and directories to do it manually. Using Ubuntu server, MadMax and docker. |
Beta Was this translation helpful? Give feedback.
-
#60 goes over why this probably won't happen. |
Beta Was this translation helpful? Give feedback.
-
aufs with round-robin and low space control is perfect for this |
Beta Was this translation helpful? Give feedback.
-
As I noticed -d flag adds copying finished plot to destination. This process is in the background (correct me if I'm wrong) so process did not wait and start plotting next plot. It is good for network location and save some extra minutes.
Is it possible to add more destination to feel more than one disk and some kind of space left checking?
Beta Was this translation helpful? Give feedback.
All reactions