Plotman plots into all destination drives regardless of free space available in a drive
See original GitHub issueDescribe the bug I have a script to fill 2 external HDs at a time. When both were around 90% full I stopped the process, connected another 5 HDs (all empty), and switched to plotman (v 0.3.1). I listed all 7 HDs as dst in the config file. After a while both of the first two HDs had 0% free space left while the other 5 had had free space down in similar amounts which is expected. In each of the 2 HDs that are 100% I found a partially constructed plot which had to be deleted I suppose. Then I used plotman to kill the remaining plotting processes still with those two HDs as destination drives (there were a total of 4 that were killed like this).
To Reproduce
Steps to reproduce the behavior, e.g.:
- Set up config with all 7 HDs as dst drives. 2 were 90% full, 5 were 1% full.
- Run other programs ‘…’
- Run Plotman with plot
- See error
Expected behavior Plotman is expected to load balance between the dst drives, but why would it continue writing to a dst drive if it doesn’t have enough free space to hold a completed plot?
System setup:
- OS: Ubuntu Server 20.04 LTS
- Method of archiving: None. Section commented out in config file.
Config
full configuration
# paste your complete configuration file contents here.
Additional context & screenshots
Issue Analytics
- State:
- Created 2 years ago
- Reactions:1
- Comments:28
Top GitHub Comments
dst is meant as a buffer. Archive is meant to distribute plots to their final resting place. Presently, it is annoying to setup archiving for local drives since you have to use rsyncd regardless. There’s a fix for that but we need a bit more testing and documentation work. It will be in the next release.
Because dst is a buffer. It is meant to be continuously cleared by archiving, or your own alternative to archiving. As such, the free space available when a plot starts is expected to have no relevance to the amount of free space when the plot ends hours later.
Available space is being checked for archiving disks not dst. See an explanation of archiving, including why, in the archiving wiki page.