Comparisons

Apache Doris vs Elasticsearch

Elasticsearch and Apache Doris are both popular in observability, cybersecurity, and real-time analytics. However, Elasticsearch can be costly in terms of storage and write resources. Apache Doris reduces these costs through efficient storage and high compression, and offers comprehensive analytical capabilities, such as JOIN and superior query performance.

Featured Migration Cases

“By replacing Elasticsearch with the Doris Commercial Distributed Version supported by VeloDB, GuanceDB showcases a big stride in improving data processing speed and reducing costs.”

70%

Cost reduction

2~3x

Faster full-text search performance

VARIANT (Data Type)

Flexible to handle semi-structured data in log tracing

“Previously, we used multiple components for complex security analysis... Adopting Doris as a unified solution has significantly improved data writes, query performance and storage efficiency.”

Faster write speeds

Better query performance

50%

Storage space savings

“Compared to the original OLAP database, query performance has improved 5-10 times, concurrency has doubled, and analysis time has dropped from 10 minutes to under 1 minute for 90% of cases, all while using just one-third of the original resources.”

Increasing report analysis concurrency

65%

Storage space reduction

SQL

Simplified query with standard SQL

Why Choose Apache Doris

	Apache Doris	Elasticsearch
Open Source License	Licensed under Apache License 2.0 Stable License since governed by the Apache Software Foundation	License changed from Apache License 2.0 to Elastic License, then to AGPL License Changing license since governed by Elastic NV
Architecture	Higher flexibility and elasticity: Strict workload isolation by workload group, powered by Linux CGroups, ideal for multi-tenancy Compute-Storage decoupled and coupled modes	Traditional deployment with limited elasticity: Soft Workload Isolation by Thread Group Does not support decoupling compute and storage
Deployment	Supports three deployment options: Cloud-native services on AWS, Azure, and GCP, as well as SaaS and BYOC versions, supported by VeloDB ( a commercial company founded by Apache Doris creators) On-premise deployment, with extended long-term support from VeloDB	Supports only two deployment options: Commercial distributed package only supports Cloud SaaS and on-premise
Real-Time Data Writes	High throughput: Indexing only on one replica Pull-based ingestion via Kafka CDC, easier and simpler Support Logstash and Beats output plugin	Low throughput: Indexing for multiple data replicas Requires additional tools like Logstash and Beats for pull-based ingestion, less convenient
Real-Time Data Storage	Low storage consumption with compression rates up to 1:5 - 1:10 Unique model supports both write and read optimization (MoW & MoR), retaining 90% of write speed when data is duplicated by key Aggregation model supports strong consistency, allows aggregated data updates, and coexists with original data Flexible Schema Change to meet dynamic business needs	High storage consumption with a compression ratio of 1:1.5 Unique model only supports write optimization, with write performance loss up to 3 times The aggregation model does not allow aggregated data to be updated and does not coexist with the original data Limited support for Schema Change
Real-Time Data Queries	Lightning-Fast in various query workloads Supports multi-table JOINs and optimization for complex analysis Easy to use with standard SQL Open MySQL ecosystem	Good at point queries, but not suited for data analysis No support for multi-table JOINs or complex analysis Difficult for users due to custom DSL Proprietary Elasticsearch ecosystem

Apache Doris

Open Source License

Licensed under Apache License 2.0
Stable License since governed by the Apache Software Foundation
Architecture
Higher flexibility and elasticity:

Strict workload isolation by workload group, powered by Linux CGroups, ideal for multi-tenancy
Compute-Storage decoupled and coupled modes
Deployment
Supports three deployment options:

Cloud-native services on AWS, Azure, and GCP, as well as SaaS and BYOC versions, supported by VeloDB ( a commercial company founded by Apache Doris creators)
On-premise deployment, with extended long-term support from VeloDB
Real-Time Data Writes

High throughput: Indexing only on one replica
Pull-based ingestion via Kafka CDC, easier and simpler
Support Logstash and Beats output plugin
Real-Time Data Storage

Low storage consumption with compression rates up to 1:5 - 1:10
Unique model supports both write and read optimization (MoW & MoR), retaining 90% of write speed when data is duplicated by key
Aggregation model supports strong consistency, allows aggregated data updates, and coexists with original data
Flexible Schema Change to meet dynamic business needs
Real-Time Data Queries

Lightning-Fast in various query workloads
Supports multi-table JOINs and optimization for complex analysis
Easy to use with standard SQL
Open MySQL ecosystem

Elasticsearch

Open Source License

License changed from Apache License 2.0 to Elastic License, then to AGPL License
Changing license since governed by Elastic NV
Architecture
Traditional deployment with limited elasticity:

Soft Workload Isolation by Thread Group
Does not support decoupling compute and storage
Deployment
Supports only two deployment options:

Commercial distributed package only supports Cloud SaaS and on-premise
Real-Time Data Writes

Low throughput: Indexing for multiple data replicas
Requires additional tools like Logstash and Beats for pull-based ingestion, less convenient
Real-Time Data Storage

High storage consumption with a compression ratio of 1:1.5
Unique model only supports write optimization, with write performance loss up to 3 times
The aggregation model does not allow aggregated data to be updated and does not coexist with the original data
Limited support for Schema Change
Real-Time Data Queries

Good at point queries, but not suited for data analysis
No support for multi-table JOINs or complex analysis
Difficult for users due to custom DSL
Proprietary Elasticsearch ecosystem

Performance Comparison

Observability & Cyber Security

The HTTP Logs benchmark is an official Elasticsearch performance test designed for log storage and analysis. It uses a real-world HTTP log dataset to evaluate indexing performance, storage efficiency, and query performance.

This benchmark comprises 11 queries commonly used in log analysis scenarios, including keyword search, time range queries, aggregations, and sorting. As a result, it is highly suitable for assessing performance in observability and network security analysis contexts.

Real-Time Analytics

ClickBench is a benchmarking tool to evaluate the performance of analytical databases. It focuses on testing the performance of large, flat tables rather than complex multi-table joins. It uses real-world data from a major web analytics platform, covering typical scenarios such as clickstream analysis and structured logs.

The benchmark consists of a set of queries that test aggregation operations and single-table performance, without involving complex joins. This makes it especially useful for evaluating databases optimized for real-time analytics and large-scale data processing.

Note: These test results are archived benchmarks captured in December 2024. Current real-time comparisons are maintained at ClickBench.

Apache Doris vs Elasticsearch

Featured Migration Cases

Why Choose Apache Doris

Apache Doris

Elasticsearch

Performance Comparison

Replacing Apache Hive, Elasticsearch and PostgreSQL with Apache Doris

Creator of Talkie migrated from Loki and built a PB-scale logging system with Apache Doris

Apache Doris for log and time series data analysis in NetEase, why not Elasticsearch and InfluxDB?

Products

Resources

Company

Apache Doris