What about some integration of "amcleanup" and "amdump" ?

I imagine a parameter definining the time of day the amdump should be
finished (= the servers are available and idle for users ...).

When amdump is running and somehow noticing that the current dumps will
take more time than until XX:YYam it could decide to stop things after
the current DLE or something like that.

Something like "amcleanup -k" but triggered by amdump itself.

Sure, it should trigger a warning or so because the overall goals might
not be met when stopping things. But it would be preferable for admins
to keep up server-availability while tinkering with backup-parameters.

What do you users/devs think?

Stefan


Reply via email to