Closed Leo3418 closed 1 year ago
Thanks for the PR! While I completely agree with the fact that a failed prune
invocation as part of a snapborg backup
run should not lead to an unsuccessful result, if the fault tolerant mode is active, the suggested changes would swallow an error also if snapborg prune
was run interactively, which is not desirable (even though the output would be shown to the user). Reading through the code, it seems to me that the exception/error handling and flow is quite complicated and might need some rework anyway...
if the fault tolerant mode is active, the suggested changes would swallow an error also if
snapborg prune
was run interactively, which is not desirable (even though the output would be shown to the user).
Does this mean that the fault tolerant mode is not supposed to apply to snapborg prune
? I see...
Reading through the code, it seems to me that the exception/error handling and flow is quite complicated and might need some rework anyway...
Possibly. An easy fix to get away with this quickly would be ignoring the exception only if the snapborg command is backup
, but this could increase the code entropy, hence a rework might be better...
The fault tolerant mode would only ignore backup errors. Pruning may also result in an error when the backup destination is unavailable, and the fault tolerant mode should cover this situation too.