Scalr Patch #4: Fixing freezing servers after mysql-backup-action with BIG databases...

Problem (from: /var/log/scalarizr.log):

2011-05-18 09:20:53,144 - INFO - scalarizr.handlers - Dumping all databases
2011-05-18 09:34:15,071 - ERROR - scalarizr.handlers - [Errno 28] No space left on device
Traceback (most recent call last):
  File "/usr/lib/python2.6/site-packages/scalarizr/handlers/mysql.py", line 721, in on_Mysql_CreateBackup
    backup.add(dump_path, os.path.basename(dump_path))
  File "/usr/lib64/python2.6/tarfile.py", line 1980, in add
    self.addfile(tarinfo, f)
  File "/usr/lib64/python2.6/tarfile.py", line 2009, in addfile
    copyfileobj(fileobj, self.fileobj, tarinfo.size)
  File "/usr/lib64/python2.6/tarfile.py", line 281, in copyfileobj
    dst.write(buf)
  File "/usr/lib64/python2.6/gzip.py", line 197, in write
    self.fileobj.write( self.compress.compress(data) )
IOError: [Errno 28] No space left on device
2011-05-18 09:34:16,296 - INFO - scalarizr.messaging.p2p.producer - Message 'Mysql_CreateBackupResult' delivered (message_id: d769239e-1ff7-47ba-a0a5-05bcf05af588)

Solution:

# diff /usr/lib/python2.6/site-packages/scalarizr/handlers/mysql.py.orig /usr/lib/python2.6/site-packages/scalarizr/handlers/mysql.py:

36,37d35
<
<
214a213
> STORAGE_TMP_DIR = "mysql-tmp"
562a562
> self._tmp_dir = os.path.join(self._storage_path, STORAGE_TMP_DIR)
700c700
< backup_path = os.path.join('/tmp', backup_filename)
---
> backup_path = os.path.join(self._tmp_dir, backup_filename)
707c707
< tmpdir = tempfile.mkdtemp()
---
> tmpdir = tempfile.mkdtemp(dir=self._tmp_dir)
1295a1296
> tmp_dir = os.path.join(path, STORAGE_TMP_DIR) if path else self._tmp_dir