Reordering jobs

Is there a way to reorder the backup jobs? Its relevant when on first start up there is a few outstanding scheduled jobs and they run from top down - would like to control the order they are run (currently the 1st is very slow but less important would like at bottom…)

You could export the job you want at the end, import it, and rename the old database into the new name.

But since you filed this for a feature, not the workaround, the feature is not there yet but has been asked.

Allow to change backup list order
Re-ordering Backup Jobs on Duplicati Landing Page which details the export/import workaround. That’s responding in part to your post, but I can’t tell whether it didn’t help your need or if you hadn’t tried it yet…

Is it possible they are running in the order they were scheduled? If so, changing the order in the GUI won’t affect the run order.

Possible, though I’d like to hear what actually happens. It looks like the scheduler gets its list here which does a SELECT here which (per SQL rules) comes out in no particular order because no ordering was requested, however if the database decides to use primary key order, then maybe it’s whatever ran and rescheduled first.

Yeah - it’s interesting how many people assume a database will always return things in the same order. :slight_smile:

looks like it really does run in the order listed and not in order of schedule.

Haven’t been into CS since university days, but isn’t possible to give a SORT BY clause to a SELECT query? I’d think ordering by schedule run time makes a bit more sense than the (somewhat arbitrary and based on historical quirks) ordering of jobs in listing…

SQL ORDER BY Keyword can do this, and Time might be what to order by if “schedule run time” means the time that got missed. Any way it’s done, someone might want it differently. For example, looking at the original post here, someone might have a very slow backup running hourly at the start of an hour, and a faster important backup that runs 30 minutes into the hour. If the backups stop (sleep, whatever) shortly before slow job runs, then when backups resume, slow job runs first because its missed time was earlier.

What one might prefer is a way to set a priority on a backup so that if lots of misses need to be done, high priority ones go first. This might actually be what you have currently, except it’s hugely indirect, and I’m not sure it holds past the first job to go into the new queue (but possibly it would – I don’t know scheduler well).