On my desktop I have 3 shares, where 2 shares had the **.DS_Store ignore pattern. On the NAS I have the same shares where all the 3 shares have the **.DS_Store ignore pattern. The pattern is not synced between two nodes but filled manually. The desktop client hangs with the NAS peer node in the 100% state under the devices heading. I have “fixed” this by having both nodes the same ignore patterns. This is weird but why does this happen?
This is expected, because ignores are local, so we expect the remote side to have something (that they have chosen to ignore).
1 Like
I completely understand from a distributed-systems point of view. But it is weird that it “hangs” on the 100% while the peer node is not accepting this files due to the ignores.
It’s a known behaviour. It’s not ideal but there already exists an issue talking about this.
The thing is that the other device is seen as (like) 99.987% in sync, as it’s missing a few files we think it should have, whatever those ignored ones were. This is, as Audrius said, expected but could probably be clarified. I see two problems:
-
It says 100%, which sounds like “done”, but still “syncing” which is confusing. We might want to round down instead, so it would get stuck at 99% which is a clearer indication that something is indeed missing?
-
It says “syncing” while the device is in fact doing no such thing. This requires activity updates, or my half measure PR to detect activity. With either of those in place it would instead say “idle”.
If it had said “Idle (99%)”, would that have made more sense to begin with?
I know this is a difficult situation, as you probably don’t want to exchange the ignores between peers. In my opinion 100%, 99% is the same as still busy. Its not that bad currently but a little weird on the eye.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.