Out of Sync: The Mismatched Global State Problem in Syncthing

[amneo]

The issue is that the global state of the main node is different from the global state of the other nodes in Syncthing. Here is an example to illustrate the problem:

Suppose you have a folder called “MyFiles” on the main node and one other node. The folder is set up to synchronize automatically between the nodes. However, after a while, I notice that the global state of the main node is different from the global state of the other node. This means that the files in the “MyFiles” folder on the main node are not the same as the files in the “MyFiles” folder on the other nodes. The file count in global at slave nodes are very less as compared to main node.

In simpler terms, the synchronization process between the nodes is not working correctly, and the files on the main node are not being updated on the other nodes as they should be.

I am running Syncthing v1.23.4 on all nodes with Windows 11.

Screenshots from Main server and slave node is attached.

[amneo]

Hi,

For the 3rd time the forced synchronisation has completed with no change.

[AudriusButkevicius]

Things are not expected to match while things are still scanning, the whole point of scanning is to collect those numbers.

[amneo]

Well the sync has competed atleast two times with no change.

Now I removed the folder on the slave machine and starting from scratch. Its 2.8 Tb of data. Now long way to go.

[Nummer378 (Max)]

I’ve had this happen once when I setup a new machine and scanned large preexisting folders. Something undefined happened during this initial scan and the machine ended up with a significantly too low global state, like here.

Resetting the database on the machine with the too low global state fixed the issue.

(In my case, the scanning did actually finish)