- Newest
- Most votes
- Most comments
Hello,
Both SERIALIZABLE and SNAPSHOT isolation are types of serializable isolation levels i.e. dirty reads, non-repeatable reads, and phantom reads are prevented according to the SQL standard. Where, SNAPSHOT isolation allows higher concurrency of transactions compared to SERIALIZABLE isolation type. Ref - https://repost.aws/articles/ARLIduLsAMSgGnG1CjjQ4IUw/understanding-redshift-isolation-levels-serializable-vs-snapshot
In your case, I suspect there are multiple queries or procedures attempting to perform DDL/DML operations on a table leading to the issue. Kindly consider adding a statement to query the database isolation level using stv_db_isolation_level system table and save it in form of information so that for every Stored Procedure invocation we can see the isolation level using SVL_STORED_PROC_MESSAGES.
Further, I recommend to reach to AWS Support Engineering team for Amazon Redshift service along with below details, so that an expert can look into the transactions and queries being run on cluster to identify the reason for Serializable Isolation error with SNAPSHOT isolation level on transactions.
- Complete Stored Procedure definition
- Complete error message being observed
- QueryID / XID / Process ID of the stored procedure run
Thank you.
Relevant content
- AWS OFFICIALUpdated 3 years ago

Hi Mounika, I will raise a ticket with Support, many thanks :)