After much learning on how to read and extract logs from our kibana instance, here is the error repartition.

"googlecode": {
  "total": 2810,
  "errors": {
    "OSError(28, 'No space left on device')": 1324,
    "NotGitRepository('No git repository was found at /": 804,
    "ValueError('Failed to uncompress archive /srv/stor": 275,
    "IntegrityError('duplicate key value violates uniqu": 152,
    "FileNotFoundError(2, \"No such file or directory: '": 126,
    "FileNotFoundError(2, \"No usable temporary director": 94,
    "StorageAPIError(ConnectionError(ProtocolError('Con": 35
  }
}

It's either that:

• we send something that was not a git repository.
• archive were not successfully uncompressed
• integrity error (which is expected for now)
• disk space problem

Note that we have less errors than previously mentioned since i cover a lesser period (multiple schedulings took place).

source:

• our elasticsearch instance
• snippet scripts to read from it
• snippet script to compute from the data read

After rescheduling of thos origins (the one we can do something about), here are the remaining errors.

"googlecode": {
  "errors": {
    "NotGitRepository('No git repository was found at /": 136
  },
  "total": 136
}

Lists of origin with integrity errors and origins that are not git repositories has been computed and stored in uffizi:/srv/storage/space/lists/loader-git-disk/origin-with-errors.googlecode.tar.gz.