- Why switch to Puma?
- Configuring Puma to replace Unicorn
- Performance caveat when using Puma with Rugged
- All-in-one package-based installations.
- Helm chart-based installations.
Puma has a multi-thread architecture which uses less memory than a multi-process application server like Unicorn. On GitLab.com, we saw a 40% reduction in memory consumption.
Most Rails applications requests normally include a proportion of I/O wait time. During I/O wait time MRI Ruby will release the GVL (Global VM Lock) to other threads. Multi-threaded Puma can therefore still serve more requests than a single process.
Beginning with GitLab 13.0, Puma is the default application server. We plan to remove support for Unicorn in GitLab 14.0.
When switching to Puma, Unicorn server configuration will not carry over automatically, due to differences between the two application servers. For Omnibus-based deployments, see Configuring Puma Settings. For Helm based deployments, see the Webservice Chart documentation.
Additionally we strongly recommend that multi-node deployments configure their load balancers to utilize the readiness check due to a difference between Unicorn and Puma in how they handle connections during a restart of the service.
MRI Ruby uses a GVL. This allows MRI Ruby to be multi-threaded, but running at most on a single core. Since Rugged can use a thread for long periods of time (due to intensive I/O operations of Git access), this can starve other threads that might be processing requests. This is not a case for Unicorn or Puma running in a single thread mode, as concurrently at most one request is being processed.
We are actively working on removing Rugged usage. Even though performance without Rugged is acceptable today, in some cases it might be still beneficial to run with it.
Given the caveat of running Rugged with multi-threaded Puma, and acceptable performance of Gitaly, we are disabling Rugged usage if Puma multi-threaded is used (when Puma is configured to run with more than one thread).
This default behavior may not be the optimal configuration in some situations. If Rugged plays an important role in your deployment, we suggest you benchmark to find the optimal configuration:
The safest option is to start with single-threaded Puma. When working with Rugged, single-threaded Puma does work the same as Unicorn.
To force Rugged auto detect with multi-threaded Puma, you can use feature flags.