Files
windmill/docker-compose.yml
T
Ruben Fiszel 7590b28108 feat(sandbox): pull/extract images with crane instead of podman (#9455)
* feat(sandbox): pull/extract images with crane instead of podman (+ add to image)

The sandboxed container runtime (`# sandbox <image>`) only ever pulls + flattens an
image (nsjail does the run), so a full container engine is overkill — and podman was
never actually in any Dockerfile, so the merged feature couldn't run in the shipped
image. Switch to crane (google/go-containerregistry): a single ~25MB static binary,
no daemon/store/root/privileged.

- docker_v2.rs: crane export -> flattened rootfs tar, crane config -> OCI config,
  crane digest -> content-addressed rootfs+config cache (cross-job dedup + automatic
  freshness), crane manifest -> pre-download size guard. DOCKER_CONFIG authfile dir.
  Cache eviction prunes the rootfs-tar cache by mtime (LRU). Pull policy honored via a
  ref->digest cache (missing/never reuse without a registry hit).
- Dockerfile + docker/DockerfileSlim{,Ee}: install the crane binary (Full/FullEe and
  the EE image inherit it via FROM the base image).
- docs + UI text + instance-setting descriptions updated (download size is compressed;
  cache is the rootfs-tar cache).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(sandbox): address CI review — digest-pinned fetch, size cap on every job, eviction race

Codex P1s:
- Fetch by the resolved digest (name@digest), not the mutable tag, so content can't
  diverge from the digest the cache is keyed under if a tag moves mid-fetch.
- Enforce the size cap on EVERY job via a cached {digest}.size sidecar (no registry call
  on cache reuse), so lowering the limit rejects already-cached oversized images.
- Eviction race: hardlink the cache tar into the job dir before tar -xf (pins the inode
  against concurrent eviction) and re-fetch if it was evicted first.
Claude P2s: atomic config sidecar (tmp+rename) + tolerate torn parse; soften the LRU
comment (mtime = creation order); sweep orphaned *.tmp.* and .size on eviction.
+digest_key/ref_key unit tests.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(sandbox): P1 cross-fs cache staging (EXDEV), Dockerfile arch fail-fast

CI re-review (Claude + Codex P1): the eviction-race hardlink crosses filesystems in the
shipped deployments — the cache is its own volume (/tmp/windmill/cache) while the job dir
is on the container fs — so hard_link returns EXDEV (not NotFound) and every sandbox job
fails. Fall back to tokio::fs::copy on a non-NotFound link error; copy reads through the
source inode so it still survives a concurrent eviction.
Also: Dockerfiles fail fast with a clear error on an unsupported arch instead of building
a 404 crane URL; ref->digest file written via tmp+rename (no torn read under missing/never).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* docs(sandbox): say 'oldest by creation time' not 'LRU' for cache eviction

Codex P2: the code evicts by tar creation time (cache hits don't touch mtime), so the
user-facing docs + instance-setting text shouldn't claim true LRU.

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-05 13:52:55 +00:00

218 lines
7.2 KiB
YAML

version: "3.7"
x-logging: &default-logging
driver: "json-file"
options:
max-size: "${LOG_MAX_SIZE:-20m}"
max-file: "${LOG_MAX_FILE:-10}"
compress: "true"
services:
db:
deploy:
# To use an external database, set replicas to 0 and set DATABASE_URL to the external database url in the .env file
replicas: 1
image: postgres:16
shm_size: 1g
restart: unless-stopped
volumes:
- db_data:/var/lib/postgresql/data
expose:
- 5432
environment:
POSTGRES_PASSWORD: changeme
POSTGRES_DB: windmill
healthcheck:
test: ["CMD-SHELL", "pg_isready -U postgres"]
interval: 10s
timeout: 5s
retries: 5
logging: *default-logging
windmill_server:
image: ${WM_IMAGE}
pull_policy: always
deploy:
replicas: 1
restart: unless-stopped
expose:
- 8000
- 2525
environment:
- DATABASE_URL=${DATABASE_URL}
- MODE=server
depends_on:
db:
condition: service_healthy
volumes:
- worker_logs:/tmp/windmill/logs
logging: *default-logging
windmill_worker:
image: ${WM_IMAGE}
pull_policy: always
deploy:
replicas: 3
resources:
limits:
memory: 2048M
# for GB, use syntax '2Gi'
restart: unless-stopped
# Uncomment to enable PID namespace isolation (recommended for security)
# Requires privileged mode for --mount-proc flag
# See: https://www.windmill.dev/docs/advanced/security_isolation
privileged: true
environment:
- DATABASE_URL=${DATABASE_URL}
- MODE=worker
- WORKER_GROUP=default
# If running with non-root/non-windmill UID (e.g., user: "1001:1001"),
# add: - HOME=/tmp
- FAVOR_UNSHARE_PID=true
depends_on:
db:
condition: service_healthy
# to mount the worker folder to debug, KEEP_JOB_DIR=true and mount /tmp/windmill
volumes:
- worker_dependency_cache:/tmp/windmill/cache
- worker_logs:/tmp/windmill/logs
## Sandboxed containers (`# sandbox <image>`) run daemonless via crane + nsjail
## inside the worker itself — no Docker socket or dind sidecar required.
## For the legacy full-compat docker (a bare `# docker`, trusted users only),
## mount the host Docker socket by uncommenting the line below. WARNING: this
## grants user scripts full access to the host Docker daemon (host filesystem
## access and privilege escalation) — only use it if you fully trust all users.
# - /var/run/docker.sock:/var/run/docker.sock
logging: *default-logging
## This worker is specialized for "native" jobs. Native jobs run in-process and thus are much more lightweight than other jobs
windmill_worker_native:
# Use ghcr.io/windmill-labs/windmill-ee:main for the ee
image: ${WM_IMAGE}
pull_policy: always
deploy:
replicas: 1
resources:
limits:
memory: 2048M
# for GB, use syntax '2Gi'
restart: unless-stopped
# Uncomment to enable PID namespace isolation (recommended for security)
# Requires privileged mode for --mount-proc flag
# See: https://www.windmill.dev/docs/advanced/security_isolation
environment:
- DATABASE_URL=${DATABASE_URL}
- MODE=worker
- WORKER_GROUP=native
- NATIVE_MODE=true
- SLEEP_QUEUE=200
depends_on:
db:
condition: service_healthy
volumes:
- worker_logs:/tmp/windmill/logs
logging: *default-logging
# This worker is specialized for reports or scraping jobs. It is assigned the "reports" worker group which has an init script that installs chromium and can be targeted by using the "chromium" worker tag.
# windmill_worker_reports:
# image: ${WM_IMAGE}
# pull_policy: always
# deploy:
# replicas: 1
# resources:
# limits:
# memory: 2048M
# # for GB, use syntax '2Gi'
# restart: unless-stopped
# # Uncomment to enable PID namespace isolation (recommended for security)
# # Requires privileged mode for --mount-proc flag
# # See: https://www.windmill.dev/docs/advanced/security_isolation
# privileged: true
# environment:
# - DATABASE_URL=${DATABASE_URL}
# - MODE=worker
# - WORKER_GROUP=reports
# - FAVOR_UNSHARE_PID=true
# depends_on:
# db:
# condition: service_healthy
# # to mount the worker folder to debug, KEEP_JOB_DIR=true and mount /tmp/windmill
# volumes:
# - worker_dependency_cache:/tmp/windmill/cache
# - worker_logs:/tmp/windmill/logs
# The indexer powers full-text job and log search, an EE feature.
windmill_indexer:
image: ${WM_IMAGE}
pull_policy: always
deploy:
replicas: 0 # set to 1 to enable full-text job and log search
restart: unless-stopped
expose:
- 8002
environment:
- PORT=8002
- DATABASE_URL=${DATABASE_URL}
- MODE=indexer
depends_on:
db:
condition: service_healthy
volumes:
- windmill_index:/tmp/windmill/search
- worker_logs:/tmp/windmill/logs
logging: *default-logging
# Combined extra services: LSP, Multiplayer, and Debugger
# Each service can be enabled/disabled via environment variables:
# - ENABLE_LSP=true (default) - Language Server Protocol for code intelligence
# - ENABLE_MULTIPLAYER=false - Real-time collaboration (Enterprise Edition)
# - ENABLE_DEBUGGER=false - Interactive debugging via DAP WebSocket
windmill_extra:
image: ghcr.io/windmill-labs/windmill-extra:latest
pull_policy: always
restart: unless-stopped
expose:
- 3001 # LSP
- 3002 # Multiplayer
- 3003 # Debugger
environment:
- ENABLE_LSP=true
- ENABLE_MULTIPLAYER=false # Set to true to enable multiplayer (Enterprise Edition)
- ENABLE_DEBUGGER=true # Set to true to enable debugger
- DEBUGGER_PORT=3003 # Debugger service port
- ENABLE_NSJAIL=false # Set to true for nsjail sandboxing (requires privileged: true)
- REQUIRE_SIGNED_DEBUG_REQUESTS=false # Set to true to require JWT tokens for debug sessions
- WINDMILL_BASE_URL=http://windmill_server:8000
volumes:
- lsp_cache:/pyls/.cache
logging: *default-logging
caddy:
image: ghcr.io/windmill-labs/caddy-l4:latest
restart: unless-stopped
# Configure the mounted Caddyfile and the exposed ports or use another reverse proxy if needed
volumes:
- ./Caddyfile:/etc/caddy/Caddyfile
- caddy_data:/data
# - ./certs:/certs # Provide custom certificate files like cert.pem and key.pem to enable HTTPS - See the corresponding section in the Caddyfile
ports:
# To change the exposed port, simply change 80:80 to <desired_port>:80. No other changes needed
- 80:80
- 25:25
# - 443:443 # Uncomment to enable HTTPS handling by Caddy
environment:
- BASE_URL=":80"
# - BASE_URL=":443" # uncomment and comment line above to enable HTTPS via custom certificate and key files
# - BASE_URL=mydomain.com # Uncomment and comment line above to enable HTTPS handling by Caddy
logging: *default-logging
volumes:
db_data: null
worker_dependency_cache: null
worker_logs: null
worker_memory: null
windmill_index: null
lsp_cache: null
caddy_data: null