disabled local data fetch for CTAS & limited local data fetch to 100k…

[mxmarg]

See description in Dremio-internal JIRA ticket: https://dremio.atlassian.net/browse/DX-86623

[mxmarg]

Hi @awaschick, @Conq1, @ravjotbrar,
I encountered a similar issue to the one discussed here (#176), when I tried to run incremental materializations for large tables and dbt was taking hours trying to fetch the records with 100 row GET requests.
I proposed a fix to disable data fetch for CTAS queries and generally limit row fetching to the same value that Dremio's UI does (100k).
Any thoughts or criticisms on this (fairly blunt) approach?

[Conq1]

I went the way of controlling it myself:
{% call statement('create_staging_table', fetch_result=False) -%}
{{ create_table_as(False, source_staging_relation, sql_select) }}
{%- endcall %}

But might be a good idea to force it for people less familiar with what's going on.

[ravjotbrar]

@mxmarg Looking into this. Agree with changing the row limit size to be 500. It looks like we might be able to make fetch false by default in the code. The only time we want to fetch is when we are getting the columns of a table.