Phase 2 — Backfill in batches (via Eventing/app code or chunked SQL): Update existing rows in batches — not in a single UPDATE that would lock the entire table. DO $$ DECLARE batch_size INT := 10000; BEGIN LOOP UPDATE "order" SET status_v2 = status WHERE status_v2 IS NULL LIMIT batch_size; EXIT WHEN NOT FOUND; PERFORM pg_sleep(0.01); -- brief pause to avoid lock contention END LOOP; END $$; Phase 3 — Contract (after all code using old column is retired): Add NOT NULL constraint, drop the old column, rename if needed. By this point no active code references the old column. For index creation on large tables in PostgreSQL, always use CREATE INDEX CONCURRENTLY in a changeSet with runInTransaction="false" to build the index without a full table lock."> Phase 2 — Backfill in batches (via Eventing/app code or chunked SQL): Update existing rows in batches — not in a single UPDATE that would lock the entire table. DO $$ DECLARE batch_size INT := 10000; BEGIN LOOP UPDATE "order" SET status_v2 = status WHERE status_v2 IS NULL LIMIT batch_size; EXIT WHEN NOT FOUND; PERFORM pg_sleep(0.01); -- brief pause to avoid lock contention END LOOP; END $$; Phase 3 — Contract (after all code using old column is retired): Add NOT NULL constraint, drop the old column, rename if needed. By this point no active code references the old column. For index creation on large tables in PostgreSQL, always use CREATE INDEX CONCURRENTLY in a changeSet with runInTransaction="false" to build the index without a full table lock." />

Prev Next

Database / Liquibase interview questions

How do you handle large table migrations in Liquibase without causing downtime?

Large table migrations — adding a non-nullable column to a 500 million row table, back-filling data, adding an index on a frequently queried column — are among the most challenging database operations to perform without application downtime. The raw DDL that Liquibase would generate often acquires table locks that block reads and writes for the duration.

The key principle is the expand/contract pattern (also called multi-phase migration or online schema change):

Phase 1 — Expand (backward compatible): Add the new column as nullable with no default, or create the new index concurrently. At this point, both old and new code can coexist with the schema.

<changeSet id="phase1-add-nullable-status" author="jane">
    <addColumn tableName="order">
        <column name="status_v2" type="VARCHAR(30)"/>  <!-- nullable initially -->
    </addColumn>
</changeSet>

Phase 2 — Backfill in batches (via Eventing/app code or chunked SQL): Update existing rows in batches — not in a single UPDATE that would lock the entire table.

<changeSet id="phase2-backfill-status" author="jane" runInTransaction="false">
    <sql>
        DO $$
        DECLARE batch_size INT := 10000;
        BEGIN
            LOOP
                UPDATE "order" SET status_v2 = status WHERE status_v2 IS NULL
                LIMIT batch_size;
                EXIT WHEN NOT FOUND;
                PERFORM pg_sleep(0.01); -- brief pause to avoid lock contention
            END LOOP;
        END $$;
    </sql>
</changeSet>

Phase 3 — Contract (after all code using old column is retired): Add NOT NULL constraint, drop the old column, rename if needed. By this point no active code references the old column.

For index creation on large tables in PostgreSQL, always use CREATE INDEX CONCURRENTLY in a changeSet with runInTransaction="false" to build the index without a full table lock.

Why should you avoid a single UPDATE statement to back-fill a new column on a 500 million row table in one Liquibase changeSet?
What runInTransaction setting is required when using CREATE INDEX CONCURRENTLY in PostgreSQL?

Invest now in Acorns!!! 🚀 Join Acorns and get your $5 bonus!

Invest now in Acorns!!! 🚀
Join Acorns and get your $5 bonus!

Earn passively and while sleeping

Acorns is a micro-investing app that automatically invests your "spare change" from daily purchases into diversified, expert-built portfolios of ETFs. It is designed for beginners, allowing you to start investing with as little as $5. The service automates saving and investing. Disclosure: I may receive a referral bonus.

Invest now!!! Get Free equity stock (US, UK only)!

Use Robinhood app to invest in stocks. It is safe and secure. Use the Referral link to claim your free stock when you sign up!.

The Robinhood app makes it easy to trade stocks, crypto and more.


Webull! Receive free stock by signing up using the link: Webull signup.

More Related questions...

What is Liquibase and what problem does it solve? What is a changeLog in Liquibase? What is a changeSet in Liquibase and how is it identified? What are the DATABASECHANGELOG and DATABASECHANGELOGLOCK tables? What is the difference between Liquibase and Flyway? What are contexts in Liquibase and how do you use them? What are labels in Liquibase and how do they differ from contexts? How does rollback work in Liquibase? What is a Liquibase precondition and why would you use one? How do you integrate Liquibase with Spring Boot? What are Liquibase change types and can you name common ones? What is the Liquibase diff command and when is it useful? What are Liquibase tags and how do you use rollback to a tag? What is runOnChange and runAlways in a Liquibase changeSet? What is the markNextChangeSetRan and markNextChangeSetRanSQL command used for? How do you manage Liquibase changeLog files in a multi-module Maven project? What is the Liquibase updateSQL command and when is it used? What is a Liquibase snapshot and what commands use it? What is the failOnError attribute in a changeSet and when should you set it to false? How does Liquibase handle transactions during migration? What is the Liquibase Hub and what does it provide? How do you use Liquibase with multiple databases or schemas in a single application? What is the Liquibase Pro checksum validity and how do clearCheckSums and validChecksums work? What are Liquibase custom change types and how do you create one? How do you handle large table migrations in Liquibase without causing downtime? What is the Liquibase Maven plugin and how do you run migrations with it? What is the Liquibase status command and how does it differ from validate? What are Liquibase property substitution and how does it work? How do you write Liquibase changeLogs in YAML format? What is the Liquibase formatted SQL changeLog format? What is the Liquibase updateTestingRollback command? How do you use Liquibase with Docker and Testcontainers in integration tests? What is the difference between addColumn with a default value and a nullable column in Liquibase for large tables? What is liquibase.properties and what configuration belongs in it?
Show more question and Answers...

MuleESB

Comments & Discussions