Closed jimklimov closed 3 months ago
hmmm sorry ... this is getting a bit complex now aren't you changeing the same function as in the other PR ?
aren't you changeing the same function as in the other PR ?
As noted in the other PR at https://github.com/oetiker/znapzend/pull/625/#issuecomment-1893231538 - indeed, same function but for a different purpose. The change here merely skips sending of datasets that are "not-enabled" by whatever definition of that is in the codebase. That other PR changes the definition.
Also both of these PRs include and supersede the less impactful changes proposed in #623 (at least, as of when it did not include numerous merge-backs from master-branch updates).
Follows up from ideas in #623
UPDATE: There was a thought about using
oracleMode
to fall back from a recursive snapshot send to one-by-one processing in case we have exclusions (and/or send recursively only some smaller sub-trees that are safe to send - no exclusions there), but looking at this code structure I do not see a recursive send in fact. Maybe it is hidden deeper in$self->zZFS->...
?..