Deleting slaves from the repository does not prevent them from coming right back the next time the LaunchAllSlaves command is sent to the host machine. So then what’s the point of deleting a slave from the repository? How do you actually remove a slave so it stays gone? Do you really have to manually delete files from the machine’s local “slaves” directory? If so, can this process please be fixed/streamlined?
Removing local slave instances:
thinkboxsoftware.com/deadlin … ing_Slaves
Disabling the ability to run multiple slave instances on a machine:
thinkboxsoftware.com/deadlin … ple_Slaves
Cheers,
Ryan
Thanks Ryan. This is a pretty confusing situation. Can I suggest renaming the Delete Slave option to something like Reset Slave Settings, since that’s effectively all it does?
Slave instances are local to the client machines. The reason for this is that you can configure your local slave instances once, and then you’ll be able to use those same instances across many repositories. The key thing to keep in mind is that the local instances aren’t tied to a particular Repository, which is why deleting from the Repository does not remove the local slave instance. It’s the same case where if a machine only has the default, single slave instance configured for it. You delete it from the repository, but the next time you run that slave, it will show up again.
The Remote Control option to delete the local slave instance works around this by connecting to the remote machine and deleting those local instances that way.
I know it’s a bit confusing, but we’re not going to rename the Delete Slave option, because you are deleting it from the Repository.
Cheers,
Ryan
Fair enough.