Comment 130 for bug 1628580

I'm also experiencing db syncs that don't fail at all but take longer than the puppet default timeout of 300 seconds (granted this is on an all-in-one VM setup with multiple controllers). Increasing this timeout by adding "timeout => N" to each exec and at least making it configurable would allow this to be fixed. As the migration scripts for components get longer and longer, and projects don't prune their early migration scripts (e.g. like the Nova 216 havana script), this time will get longer and longer and impact more kinds of environments. The fix here that does a "retry" will probably alleviate this to some degree as if you retry a migration script that only got through half its migrations, it will pick up where it left off, but even better would be to allow the script to complete in one run. Would tripleo be open to additional gerrits in this regard?