Why Docker Struggles with MySQL: 7 Critical Reasons

Containers are designed to ensure software runs consistently across different environments, but the question remains: should MySQL be containerized?

Many advocate containerizing MySQL solely for the benefits of containers, while opponents cite performance and data‑security concerns. Below are the main reasons Docker is not ideal for MySQL.

Data Security Issues

Storing data inside a container is discouraged because containers can be stopped or removed at any time, causing data loss. Although volumes can persist data, Docker’s volume design around Union‑FS offers limited guarantees, and sudden crashes may corrupt the database.

Performance Issues

MySQL, as a relational database, has high I/O demands. Running multiple instances on a single physical host can create I/O bottlenecks, dramatically reducing read/write performance.

During a Docker‑focused discussion, an architect from a state‑owned bank noted that database performance bottlenecks usually stem from I/O, and Docker’s model of multiple containers sharing storage can exacerbate this problem.

Several strategies can mitigate these issues:

1. Separate Database Engine from Data

When using Docker for MySQL, keep the database program in the container while storing data on shared storage outside the container. Avoid placing data on the host’s filesystem to reduce impact on the host.

2. Use Lightweight or Distributed Databases

Deploying lightweight or distributed databases in Docker aligns with Docker’s philosophy of automatically restarting failed services.

3. Proper Application Placement

For I/O‑intensive workloads, deploying the database on a physical machine or KVM is preferable; many cloud providers run their flagship databases directly on bare metal.

State Management Issues

Docker’s horizontal scaling is intended for stateless services, not databases. Stateful services require separate storage solutions, and Docker’s fast scaling does not apply to databases.

Both Tencent Cloud’s TDSQL and Alibaba Cloud’s OceanBase run on physical machines rather than Docker.

Resource Isolation Concerns

Docker relies on cgroups for resource limits, which can cap consumption but cannot fully isolate resources from other processes. Excessive resource usage by other applications can degrade MySQL’s performance.

Higher isolation levels increase overhead, and while Docker excels at horizontal scaling for stateless services, databases do not fit this model.

Can MySQL Run in Containers?

MySQL is not completely unsuitable for containerization. Viable scenarios include:

Workloads tolerant of data loss (e.g., user search indexing) can shard databases to increase throughput.

Running lightweight or distributed databases where Docker can automatically restart failed containers.

Using middleware and orchestration to provide auto‑scaling, disaster recovery, and multi‑node capabilities.

Typical cases such as Tongcheng Travel, JD.com, and Alibaba have successfully containerized their databases.

Source: https://www.toutiao.com/article/6675622107390411276/ (© original author)