This is a continuation of http://github.com/holland-backup/holland/issues/closed#issue/11 as the original issue has been resolved.
The problem I now have is the following:
Tested and I think this bug itself may be resolved. I was able to generate a new mysqldump configuration file using mk-config (which previously did not work), but now I get this when trying to run a backup:
[root@holland ~]# holland bk default
Holland 0.9.9 started with pid 6907
-----> Starting backup run <----
Acquired backup lock.
Loading config for backupset default
Loading plugin mysqldump
Prepared backup spool /var/spool/holland/default/20100426_091228
Loading /root/.my.cnf [/root/.my.cnf]
Loading ~/.my.cnf [/root/.my.cnf]
Unexpected exception caught. This is probably a bug. Please report to the holland development team.
Traceback (most recent call last):
File "/usr/lib/python2.4/site-packages/holland/commands/backup.py", line 101, in run_backup
backup(jobname, dry_run)
File "/usr/lib/python2.4/site-packages/holland/core/backup.py", line 126, in backup
estimated_size = plugin.estimate_backup_size()
File "/usr/lib/python2.4/site-packages/holland/backup/mysqldump/plugin.py", line 93, in estimate_backup_size
self.schema.refresh(db_iter=db_iter, tbl_iter=tbl_iter)
File "/usr/lib/python2.4/site-packages/holland/lib/mysql/schema/base.py", line 112, in refresh
for table in tbl_iter(database.name):
File "/usr/lib/python2.4/site-packages/holland/lib/mysql/schema/base.py", line 249, in __call__
for metadata in self.client.show_table_metadata(database):
File "/usr/lib/python2.4/site-packages/holland/lib/mysql/client/base.py", line 174, in show_table_metadata
return self._show_table_metadata50(database)
File "/usr/lib/python2.4/site-packages/holland/lib/mysql/client/base.py", line 124, in _show_table_metadata50
row['is_transactional'] = row['engine'].lower() in ['innodb']
AttributeError: 'NoneType' object has no attribute 'lower'
One or more backupsets failed to run successfully
This backup run of 1 backupset took 0.82 seconds
-----> Ending backup run <----
A fix for this has been pushed, but now I end up with:
[root@holland noarch]# holland bk default
Holland 0.9.9 started with pid 11249
-----> Starting backup run <----
Acquired backup lock.
Loading config for backupset default
Loading plugin mysqldump
Prepared backup spool /var/spool/holland/default/20100426_102107
Loading /root/.my.cnf [/root/.my.cnf]
Loading ~/.my.cnf [/root/.my.cnf]
Unexpected exception caught. This is probably a bug. Please report to the holland development team.
Traceback (most recent call last):
File "/usr/lib/python2.4/site-packages/holland/commands/backup.py", line 101, in run_backup
backup(jobname, dry_run)
File "/usr/lib/python2.4/site-packages/holland/core/backup.py", line 126, in backup
estimated_size = plugin.estimate_backup_size()
File "/usr/lib/python2.4/site-packages/holland/backup/mysqldump/plugin.py", line 94, in estimate_backup_size
return sum([db.size for db in self.schema.databases])
File "/usr/lib/python2.4/site-packages/holland/lib/mysql/schema/base.py", line 163, in size
return sum([table.size for table in self.tables if not table.excluded])
File "/usr/lib/python2.4/site-packages/holland/lib/mysql/schema/base.py", line 200, in size
return self.data_size + self.index_size
TypeError: unsupported operand type(s) for +: 'NoneType' and 'NoneType'
One or more backupsets failed to run successfully
This backup run of 1 backupset took 0.75 seconds
-----> Ending backup run <----
These appear to be possibly related since they both involve the estimate_backup_size function. Andy mentioned the first error was caused by VIEWs so I suspect this may be the case here as ewll.
Test DBs are:
mysql> show databases;
+--------------------+
| Database |
+--------------------+
| information_schema |
| employees |
| menagerie |
| mysql |
| sakila |
| test |
| world |
+--------------------+
7 rows in set (0.00 sec)