Unclear Failure

[loe]

I have tried git-sweep on a few of my larger repositories and I get an error but I'm not sure how to progress.

I see:
$ git sweep preview
Fetching from remote
len([huge array of stuff]) != len([another huge array])

and exit status is 1.

[robmadole]

If you run git-sweep preview --nofetch do you still get an error?

Also, what about just running git fetch? Does that do anything strange?

Is your remote called "origin" or something else?

[loe]

Running git sweep preview --nofetch seems to just spin my fan. Its been running for a few minutes now with no output.

git fetch doesn't output anything exits 0.

[loe]

Traceback (most recent call last): File "/usr/local/share/python/git-sweep", line 9, in <module> load_entry_point('git-sweep==0.1.1', 'console_scripts', 'git-sweep')() File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/gitsweep/entrypoints.py", line 9, in main CommandLine(sys.argv).run() File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/gitsweep/cli.py", line 89, in run self._sweep() File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/gitsweep/cli.py", line 126, in _sweep ok_to_delete = inspector.merged_refs(skip=skips) File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/gitsweep/inspector.py", line 37, in merged_refs with_extended_output=True, with_exceptions=False) File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/git/cmd.py", line 346, in execute stdout_value, stderr_value = proc.communicate() File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/subprocess.py", line 751, in communicate return self._communicate(input) File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/subprocess.py", line 1291, in _communicate stdout, stderr = self._communicate_with_poll(input) File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/subprocess.py", line 1345, in _communicate_with_poll ready = poller.poll() File "/usr/local/Cellar/python/2.7.2/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/async/__init__.py", line 21, in thread_interrupt_handler prev_handler(signum, frame) KeyboardInterrupt

Thats the output on Ctrl+C

[robmadole]

Looks like the Git library we use is just spinning on the commands to find the merge branches. How many branches do you have? Is it an un-holy amount?

[loe]

991, 649, 364 on the different applications (on remote) I have fewer locally because I haven't touched every bug.

[robmadole]

So that's 3 repositories? repo1 has 991, repo2 has 649, and repo3 has 364?

That's quite a few. It will launch a sub-process for each remote branch and run a git command (so it's doing that 991 times for repo1). I think we are just seeing the compound effects of that. I'm sorry to say that I don't have a quick fix for this, but I'll give it some thought. Maybe run in batches or to at least have a progress bar to let you know it's working on something.

Since it was spinning up your fan, do you think it was pegging your CPU? (no need to check it again, just curious)

[loe]

Git did go to 100% CPU then die and I think python was hung in I/O wait.

[taybin]

I have this issue too. We have 450 branches. --nofetch prevents the len() != len() problem.

It still takes awhile, but that's fine.

[loe]

--nofetch is working for us now, just cleaned out a couple hundred!

[robmadole]

Cool, sounds like this is less urgent but still a problem. Thanks for letting me know. It's on the list for the next release.

[mwarkentin]

This happened for one of our repos as well.. trying it with --nofetch now.