Docker & Kubernetes in one dashboard
Tail logs from containers and pods, check host stats, manage workloads, and route alerts to your team chat. Set RUNTIME_MODE to docker, kubernetes, or both. No Elasticsearch cluster required.
Kubernetes hub
Namespace picker, pods, deployments, HPAs, services, cluster events, pod logs, shell, RBAC patterns, and event-based alerts. Mount kubeconfig or run in-cluster with a ServiceAccount.
Live log streams
stdout/stderr over WebSockets from Docker containers or Kubernetes pods. No polling delay.
Container and pod RBAC
Wildcard or regex visibility rules for containers, namespaces, and pods plus two-layer action rights: server ALLOW_* gates and per-user can_* flags.
Kubernetes hub
Namespace-scoped overview with pods, deployments, HPAs, services, and cluster events. Unified dashboard when RUNTIME_MODE=both.
CPU and memory
Host, container, and pod stats with live updates. About 30 days of history in SQLite.
Small footprint
Go backend, ~30MB RAM typical. One container image for Docker and Kubernetes.
Optional database
No-auth without DB_PATH runs in-memory. Mount a volume when you want users and audits to survive restarts.
Workload controls
Start, stop, restart, remove, and optional in-browser shell for containers and pods when both server and user permissions allow. Actions are audit-logged in auth mode.
Container and pod detail views
Inspect metadata, ports, labels, health state, linked services, and live logs from dedicated detail pages.
Built-in CLI
docklog reset-password, config, version, and agent subcommands ship in the Docker image on PATH.
Admin notifications
Route workload, security, health, admin, and intelligent alert events to Slack, Microsoft Teams, Discord, or any HTTPS webhook.
Intelligent alert rules
Monitor logs, Docker events, Kubernetes events, and CPU or memory metrics with per-rule severity, scope, throttling, and delivery to your configured channels.
Where people use it
Docker hosts, homelabs, shared VPS setups, and clusters where teams need scoped log access.
Prod and staging on one host
Give each developer container or namespace patterns so they tail their own services without seeing prod databases or unrelated stacks.
Small team, shared VPS
Auth mode, RBAC rules, audit log. Everyone gets log access without handing out SSH keys to the whole box.
Docker and Kubernetes together
RUNTIME_MODE=both on one instance: Docker socket for local containers and kubeconfig for cluster pods, with one RBAC model across both.
Homelab or NAS
One image, socket or kubeconfig mount, done. Works on Pi and small cloud instances where you do not want another service to babysit.
Client or agency work
Hand a client read-only log access to their stack without opening the full Docker daemon, cluster admin, or your internal workloads.
Try it on your server
Setup guide covers Compose, docker run, env vars, Kubernetes, and reverse proxy notes.