Assert code: 1000

Question

Hello,

We have two clusters sharing data via datashare. Since this morning, we have encountered the following error:

```
[2023-10-24 07:47:26] [XX000] ERROR:
[2023-10-24 07:47:26] -----------------------------------------------
[2023-10-24 07:47:26] error:  Assert
[2023-10-24 07:47:26] code:      1000
[2023-10-24 07:47:26] context:   current_dml_end_version < structure_change_version - current_dml_end_version=38816924, structure_change_version=38814355.
[2023-10-24 07:47:26] query:     0
[2023-10-24 07:47:26] location:  StorageOidManager.cpp:1366
[2023-10-24 07:47:26] process:   padbmaster [pid=1073848973]
[2023-10-24 07:47:26] -----------------------------------------------
```

When attempting to accessing some datashared objects. It does not affect all objects.

Why might we be encountering this issue and how can we resolve it? Thanks in advanced.

Best,
Patrick

Answer

I am experiencing similar errors in our redshift cluster, accessing tables across databases, and it also first appeared in our system on 10/24/23.
```
[Amazon](500310) Invalid operation: 
  -----------------------------------------------
  error:  Assert
  code:      1000
  context:   current_dml_end_version < structure_change_version - current_dml_end_version=3484461, structure_change_version=3474690. 
  query:     0
  location:  StorageOidManager.cpp:1366
  process:   padbmaster [pid=1073816691]
  -----------------------------------------------
```

Each time it appears, I've found that I can narrow down the scope of the issue to one particular table involved in whichever query encountered that error, and reproduce the error by EXPLAINing a simple query against the table, e.g. 
```
explain select 1 from db_name.schema_name.table_name;
```
I can fix the problem for a particular table by connecting to its host database, adding an empty column to the table, and then immediately dropping that column:
```
alter table schema_name.table_name add column _temp varchar(1) null;
alter table schema_name.table_name drop column _temp;
```
However, the issue keeps recurring, disrupting our nightly data warehouse build.  Each time it's a different table, and we have to manually perform the fix above. I'm guessing it's an issue with the synchronization of the table structure between the databases - hence the "version" mismatch in the error message - but that's just a guess.

I would also like to know the root cause of these errors and how to prevent them from occurring.

Assert code: 1000

Contenuto pertinente