#28 Automatically retry backups for n times

Open
opened 2 years ago by import_bot · 2 comments

To handle connection issues (such as SSH’s “broken pipe”, which you cannot always prevent) you might want to add an option to retry backups (for n times) when they fail.

This is what I’ve also explained here: https://github.com/borgbackup/borg/issues/636#issuecomment-257319412


Imported from Taiga issue 27 (to do). Created on 2016-10-31T15:55:36+0000 by rugk.

To handle connection issues (such as SSH's "broken pipe", which you cannot always prevent) you might want to add an option to retry backups (for n times) when they fail. This is what I've also explained here: https://github.com/borgbackup/borg/issues/636#issuecomment-257319412 --- Imported from Taiga issue 27 (to do). Created on 2016-10-31T15:55:36+0000 by rugk.

This would be good. Any indication on how this could be setup in the schema?

https://projects.torsion.org/witten/borgmatic/src/branch/master/borgmatic/config/schema.yaml

Simply adding a “num_retry” somewhere around here? https://projects.torsion.org/witten/borgmatic/src/branch/master/borgmatic/config/schema.yaml#L129

This would be good. Any indication on how this could be setup in the schema? > https://projects.torsion.org/witten/borgmatic/src/branch/master/borgmatic/config/schema.yaml Simply adding a "num_retry" somewhere around here? https://projects.torsion.org/witten/borgmatic/src/branch/master/borgmatic/config/schema.yaml#L129
witten commented 2 months ago
Owner

Yup, the storage section seems like the logical place, given precedent. I might just call it retries or retries_count though.

Variant idea: Make retrying some number of times the default behavior, and perhaps don’t even make it configurable at all; just hard-coded. Rationale: Almost everyone probably wants retrying on, and anyone who really cares about when retries are happening can go spelunking in the logs.

Thoughts? I don’t feel super strongly either way.

Yup, the `storage` section seems like the logical place, given precedent. I might just call it `retries` or `retries_count` though. Variant idea: Make retrying some number of times the default behavior, and perhaps don't even make it configurable at all; just hard-coded. Rationale: Almost everyone probably wants retrying on, and anyone who really cares about when retries are happening can go spelunking in the logs. Thoughts? I don't feel super strongly either way.
Sign in to join this conversation.
No Milestone
No Assignees
3 Participants
Due Date

No due date set.

Dependencies

This issue currently doesn't have any dependencies.

Loading…
Cancel
Save
There is no content yet.