On Oct 1, 2015, at 3:47 PM, Kern Sibbald <kern AT sibbald DOT com> wrote:

Hello Dan,

The tests will not work for your nodes that are reporting a difference in commits.  This difference is probably because at one point in the past I rewrote the git history slightly to remove some unwanted commits.

There are two ways to correct the problem on those nodes:

1. Rewind the repo and repull it with the following commands:

   cd <repo>
   git checkout Branch-xxx
   git reset --hard 1ef84681d13d0c35
   git pull

This should work on any branch ...

2. Delete the repo and reclone it.

 I have tested #1, but my Internet connection is via satellite and is so terribly slow that I cannot lookup the exact command for cloning the database.

Best regards,
Kern
  

On 15-10-01 05:28 AM, Dan Langille wrote:

I am running regression testing on both Bacula 7 and Bacula 5. Both are still in heavy use so regression testing is still important.

My Bacula 7 tree seems ok, and all nodes report:

$ git status
On branch Branch-7.2
Your branch is up-to-date with 'origin/Branch-7.2'.
nothing to commit, working directory clean


My 5.x source trees are not consistent.  Most nodes report:

$ git status
On branch Branch-5.2
Your branch is ahead of 'origin/Branch-5.2' by 1354 commits.
  (use "git push" to publish your local commits)
nothing to commit, working directory clean


A few nodes report

$ git status
On branch Branch-5.2
Your branch is up-to-date with 'origin/Branch-5.2'.
nothing to commit, working directory clean