Skip to content

Why users are migrating from ClickHouse to StarRocks

For the past decade, ClickHouse has been recognized as the leader in fast analytical databases, but times are changing. More and more users are beginning to recognize its limitations, especially for modern analytics scenarios. Some of these limitations include:
No Joined Table

ClicHouse struggles with multi-table joins, requiring denormalization upstream. This makes analytics inflexible and complex to manage.

Limited support for
data changes

ClickHouse can only handle real-time analytics as long as there are no updated records in the data stream. As the frequency of data updates increases, ClickHouse' performance begins to fluctuates and degrades substantially.

Lack of

ClickHouse provides great performance when concurrent sessions are limited. However, ClickHouse lacks the advanced capabilities required to deliver strong performance when queries become complex or concurrency is high.

to Scale

ClickHouse's single server query performance is good, but scaling a ClickHouse cluster requires manual re-balancing of data, which can be a huge headache for everyone.

Because of this, many ClickHouse users have started migrating to StarRocks. With StarRocks, these former ClickHouse users are able to enjoy significant query performance without the limitations of ClickHouse holding them back.

StarRocks vs. ClickHouse

1.7x Greater Performance In Single-Table
2.3x Greater Performance In Low-Cardinality Scenarios
#1 Class-leading multi-table JOIN performance

Free yourself from denormazlied tables

Join relationships are the foundation of modern analytics, but they also pose a challenge to query performance.

ClickHouse has tried to circumvent this challenge by focusing on single-table query performance. Because of this, ClickHouse users have to flatten normalized tables into a single flat table before data ingestion. This step not only makes the analytics inflexible, but also adds complexity to the data processing pipeline.

StarRocks' native support for distributed JOIN algorithms and its cost-based optimizer best-in-class query performance against both single and multi-table joined queries. With StarRocks, users get to keep their normalized tables and do JOIN operations on the fly, simplifying their data ingestion pipeline, improving data freshness, and cutting down on ETL costs.

Embrace mutable data

Mutable data is a common byproduct of business activities. It can be caused by glitches in the underlying data pipeline or it can simply be a part of normal business logic.

ClickHouse, like most other analytical databases, doesn't have native support for UPDATE and DELETE operations. Instead, it provides a MUTATION operation to asynchronously ALTER TABLE. ClickHouse's asynchronous MUTATION operations can cause degraded and unpredictable query latency when there are frequent data ingestions. This is not suitable for real-time analytics with data mutations.

StarRocks' primary key table is designed to handle mutated data analytics natively. Using its primary key index, data changes are resolved at data ingestion, delivering uncompromising query performance to users.


Scale your analytics seamlessly

ClickHouse's performance only lasts as long as you're doing aggregation queries. Once scenarios become more demanding, like complex, high-concurrency queries, performance starts to crater.

StarRocks has none of these limitations. With a blazing-fast execution engine and a suite of advanced features like its built-in cost-based optimizer, StarRocks accelerates queries by reusing cached and saved query results. This makes it great for scaling concurrency in any scenario.

StarRocks' materialized view (MV) works seamlessly with the supports of incremental updated (sync' MV only) and automatic query rewrite.

StarRocks' query cache can save intermediate computation results of executed queries. New queries that are (partially) semantically equivalent can reuse the cached results. 

Compare ClickHouse to StarRocks

Designed for the analytics needs of modern enterprises, StarRocks delivers the capabilities and performance. ClickHouse can't say the same.


Legacy scatter-gather architecture
Only partial SQL syntax support
Asynchronized real-time updates
Supports limited number of concurrent users
Zookeeper-based operations
    No cost based optimizer
    No distributed joins
    No data lake query support
    No support for federated queries


Modern MPP architecture
Full SQL syntax support
Synchronized real-time updates
High concurrency with 10,000+ QPS
No 3rd party dependencies
Built-in cost based optimizer
Distributed joins
Query support for Hive, Hudi, Iceberg, and Delta
Federated queries with Hive, MySQL, ES, and JDBC sources

Talk to an engineer

Have questions about CelerData and StarRocks? You can connect with our team of solutions architects and experienced engineers who can answer all of your questions and even offer a personalized demo aligned with your specific needs and analytics scenarios.