You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Document and ship a systemd unit file example for running memayu serve as a persistent service #49
Currently there's no guidance for keeping memayu serve running persistently and restarting it automatically (e.g. after a Raspberry Pi reboot). This doesn't need automation built into the binary itself — a documented, ready-to-copy systemd unit file is sufficient and is the standard expectation for this kind of self-hosted tool.
Fix
Add a contrib/systemd/memayu.service example file to the repository, and reference it from the README's self-hosted deployment section:
(Adjust paths/user to match actual installed conventions — this is illustrative, verify against how the binary is actually expected to be installed via Homebrew/npm/manual build.)
Document in the README: where to place this file, how to enable it (systemctl enable --now memayu), and how it interacts with .env (via EnvironmentFile, once Issue 1 in this batch is fixed and confirmed working) or config.toml.
Acceptance criteria
contrib/systemd/memayu.service exists in the repo with working example content.
README has a short "Running as a service" section linking to it, including enable/start/status/logs commands (systemctl status memayu, journalctl -u memayu -f).
Manually verified on a real device (e.g. the Raspberry Pi used for this testing round) that the unit file works as documented: survives a reboot, restarts on crash.
Implemented by PR #65 (docs-only, fixes #49) — ships a systemd unit (contrib/systemd/memayu.service: unprivileged memayu user, Restart=on-failure, persisted DB at /var/lib/memayu/memayu.db) and a Docker Compose example (restart: unless-stopped, readiness /api/health healthcheck, memayu-data volume) plus a new TROUBLESHOOTING.md (run memayu doctor first) and README pointers. No code/schema changes. Reviewed by g3333kbot: APPROVED (1 warning: compose healthcheck waits for "ready" so container shows unhealthy until first-run setup — should be called out in TROUBLESHOOTING.md too).
Activity
savioruz commentedon Aug 25, 2026
Implemented by PR #65 (docs-only, fixes #49) — ships a systemd unit (
contrib/systemd/memayu.service: unprivilegedmemayuuser, Restart=on-failure, persisted DB at /var/lib/memayu/memayu.db) and a Docker Compose example (restart: unless-stopped, readiness/api/healthhealthcheck,memayu-datavolume) plus a newTROUBLESHOOTING.md(runmemayu doctorfirst) and README pointers. No code/schema changes. Reviewed by g3333kbot: APPROVED (1 warning: compose healthcheck waits for "ready" so container shows unhealthy until first-run setup — should be called out in TROUBLESHOOTING.md too).chore(docs): add systemd service example, docker-compose, and trouble…