InvalidCachedStatementError during a database migration

[jcugat]

• asyncpg version: 0.18.3
• PostgreSQL version: 9.6.8
• Do you use a PostgreSQL SaaS? If so, which? Can you reproduce
  the issue with a local PostgreSQL install?: AWS RDS Aurora PostgreSQL, can reproduce it locally
• Python version: 3.6.8
• Platform: Linux
• Do you use pgbouncer?: No
• Did you install asyncpg with pip?: Yes
• If you built asyncpg locally, which version of Cython did you use?:
• Can the issue be reproduced under both asyncio and
  uvloop?: uvloop

I recently got bit by what's described at the end of this issue: #103 (review)

I did a schema migration of a table while the service was running, and a few errors like this appeared (once for each process):

    results = await connection.fetch(query)
  File "/usr/local/lib/python3.6/site-packages/asyncpg/connection.py", line 421, in fetch
    return await self._execute(query, args, 0, timeout)
  File "/usr/local/lib/python3.6/site-packages/asyncpg/connection.py", line 1414, in _execute
    query, args, limit, timeout, return_status=return_status)
  File "/usr/local/lib/python3.6/site-packages/asyncpg/connection.py", line 1422, in __execute
    return await self._do_execute(query, executor, timeout)
  File "/usr/local/lib/python3.6/site-packages/asyncpg/connection.py", line 1444, in _do_execute
    result = await executor(stmt, None)
  File "asyncpg/protocol/protocol.pyx", line 196, in bind_execute
asyncpg.exceptions.InvalidCachedStatementError: cached statement plan is invalid due to a database schema or configuration change

The problem is that the query executed was inside a transaction, so there was no way for asyncpg to recover from that state (and why the exception was raised).

Is changing the parameter statement_cache_size=0 the only way to avoid this? Do you know if there is a lot of performance penalty when disabling this cache?

[elprans]

Disabling statement cache for migrations is the way to avoid this. The performance difference in this context is negligible.

[jcugat]

We've applied this change and it's working fine 👍

[andrew222651]

Disabling statement cache for migrations is the way to avoid this. The performance difference in this context is negligible.

Is there a way to do this using SQLAlchemy?