Definition
A helper process deployed alongside a service to add capabilities without changing the service itself.
How it works
Runs in the same pod/host and handles cross-cutting work — proxying, logging, secrets, mTLS — so every service gets it for free in any language. The data plane of a service mesh is sidecars. The cost is one extra container per instance and the latency of an in-host hop.
Common questions
What is Sidecar?
A helper process deployed alongside a service to add capabilities without changing the service itself.
How does Sidecar work?
Runs in the same pod/host and handles cross-cutting work — proxying, logging, secrets, mTLS — so every service gets it for free in any language. The data plane of a service mesh is sidecars. The cost is one extra container per instance and the latency of an in-host hop.
What is Sidecar used for in system design?
Runs in the same pod/host and handles cross-cutting work — proxying, logging, secrets, mTLS — so every service gets it for free in any language. The data plane of a service mesh is sidecars. The cost is one extra container per instance and the latency of an in-host hop.