Redshift structure_change_version assert error

Question

Once in a while we seem to get below error in our redshift cluster. It is usually solved by deleting and recreating the affected table. I would however like to understand more in detail why this is happening and what we can do to prevent, since it affect our flows. What could cause this behaviour?

```
    error:  Assert
    code:      1000
    context:   current_dml_end_version < structure_change_version - current_dml_end_version=10859093, structure_change_version=10751101. 
    query:     0
    location:  StorageOidManager.cpp:1366
    process:   padbmaster [pid=1073848832]
```

Answer

Assert errors are database Oops moments. You can open a Support Ticket and they can provide more clarity and details around it.
But, as a best practice, run your Production workloads on `trailing` track as that is more stable compared to `current` track.

Redshift structure_change_version assert error

相关内容