Change log for File system repository

Under the changes tab of a configuration a new changeset is added every minute when a file system repository is checked. A changeset is added even if there are no changes. So for the last 4 weeks I have over 1900 pages of changesets.

Using Continua server 1.9.2.589.

Hi Anders,

Can you provide us with the settings for this repository? Have you specified any filtering patterns? Can you also check whether there are any errors listed on the Event Log page (under Administration) for this repository?

I see now that there is a related event log error:

File System Repository

Errors occurred while synchronising the cache folder at ‘E:\ContinuaShare\Rc\3fde9b40’ for file system repository ‘vmAC’.
Errors occurred while deleting the following folders from the repository cache:
E:\ContinuaShare\Rc\3fde9b40\6.06.0: (145) The directory is not empty: [\?\E:\ContinuaShare\Rc\3fde9b40\6.06.0]
E:\ContinuaShare\Rc\3fde9b40\6.99.0: (145) The directory is not empty: [\?\E:\ContinuaShare\Rc\3fde9b40\6.99.0]

This event has been reported 23145 times.

The file system source of this repository contains file system junctions (directory links), but they are copied as regular directories as expected, so I don’t see how that should be an issue.

I tried to reset the repository. That cleared the empty changesets, but new ones were added after the reset, as well as new errors to the event log. I also tried to manually delete the two folders from the log before resetting, but with the same result.

Hi Anders,

There does appear to be an issue with directory junction points in our comparison code for the File System repository, although it’s not the same error that you are seeing. In our test, the junction point is ignored, but not the files under the junction point, so we see an error when copying a file into a non-existent folder. We’ll work on a fix for this and investigate further.

I can see from your direct message that the file system is on a UNC path, which may account for the error occurring in a different place. We’ll adapt our tests to handle this situation and work on get a fix out for this next week.