Skip to content
Skills Library
DevOps & cloud · Automation

dbnsync-deploy

/dbnsync-deploy

Install, update, check or troubleshoot DbnSync, the Docker service that archives Databento data to S3, on a NAS (Synology, QNAP, UGREEN), a Linux server, Docker Desktop or a DigitalOcean Droplet. Use this whenever someone wants to deploy, set up, run, upgrade, move or repair DbnSync or its RustFS, Seq or Caddy containers, give readers S3 keys, or reach the archive from the internet, even if they only say "get this running on my NAS" or "the dashboard is down".

Repository
DbnSuite ↗
Licence
MIT
Last repository change
12.11.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
8
Scripts
2
HTML comments
0
URLs in SKILL.md
9
Deprecation

No signals detected. Repository changed on 12.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What to check before installing
  • 2 scripts in this package. Review them before installing, as Claude may execute them.

Scripts: scripts/check-env.sh, scripts/fill-secrets.sh

What it does and when to use it

Install, update, check or troubleshoot DbnSync, the Docker service that archives Databento data to S3, on a NAS (Synology, QNAP, UGREEN), a Linux server, Docker Desktop or a DigitalOcean Droplet. Use this whenever someone wants to deploy, set up, run, upgrade, move or repair DbnSync or its RustFS, Seq or Caddy containers, give readers S3 keys, or reach the archive from the internet, even if they only say "get this running on my NAS" or "the dashboard is down".

Triggers from the description

get this running on my NASthe dashboard is down
Original author description

Install, update, check or troubleshoot DbnSync, the Docker service that archives Databento data to S3, on a NAS (Synology, QNAP, UGREEN), a Linux server, Docker Desktop or a DigitalOcean Droplet. Use this whenever someone wants to deploy, set up, run, upgrade, move or repair DbnSync or its RustFS, Seq or Caddy containers, give readers S3 keys, or reach the archive from the internet, even if they only say "get this running on my NAS" or "the dashboard is down".

How it works

  1. Say this first
  2. Rules that protect the person
  3. 1. Find out where it is going
  4. 2. Get the deployment files
  5. 3. Make the archive folder
  6. 4. Fill in .env
  7. 5. Start it
  8. 6. Prove it works
  9. 7. Afterwards
Original SKILL.md headings · 122 lines, 1,370 words
SKILL.md preview
---
name: dbnsync-deploy
description: Install, update, check or troubleshoot DbnSync, the Docker service that archives Databento data to S3, on a NAS (Synology, QNAP, UGREEN), a Linux server, Docker Desktop or a DigitalOcean Droplet. Use this whenever someone wants to deploy, set up, run, upgrade, move or repair DbnSync or its RustFS, Seq or Caddy containers, give readers S3 keys, or reach the archive from the internet, even if they only say "get this running on my NAS" or "the dashboard is down".
---

# Deploy DbnSync

DbnSync is three containers: RustFS (S3, holding the archive), Seq (logs) and the worker, with Caddy as an optional fourth for reaching it from the internet. This skill takes someone from nothing to a healthy service, then helps them keep it that way. Be prescriptive: use the defaults unless the person has a reason not to, and don't offer a menu.

The people using this are often traders, not system administrators. Say what a step does in plain words before you do it, and explain any failure in terms of what they should do next.

Give one action at a time. When the person has to do something, ask for that one thing in a few lines, then wait for them to say it is done. Opening a file, pasting into it and saving it are three actions, not one.

## Say this first

Before anything else, tell the person, in your own words but keeping all three points:

- DbnSuite is an independent project with no official connection with Databento, Inc.
- It is for data they are licensed to receive; their Databento agreements govern storing, sharing and redistributing it.
- It is free software, provided as is, and they run it at their own risk.

## Rules that protect the person

These exist because the worker acts on a real Databento account and the archive can take days to rebuild.

- **Keep secrets out of the conversation.** Anything shown to you leaves their machine. Ask them to type `DATABENTO_API_KEY` into `.env` themselves, in an editor. Don't print `.env`, don't `grep` its values, and don't echo a secret. `scripts/check-env.sh` reports on it without showing any value.
- **Never turn on spending.** Leave `DBNSYNC_ALLOW_PAID_JOBS` and the `DBNSYNC_MAX_*` settings alone: the default of $0 is what makes this safe. If someone wants paid jobs, send them to the [spend policy docs](https://dbnsuite.squideyes.com/docs/dbnsync#spend) and have them set it themselves, knowing Databento bills their account. When jobs are refused as over the limit, the fix is the plan, `DbnSync__HistoryStart`, `DbnSync__ReachBackTo` or the schemas, never a higher limit.
- **Protect the archive.** Never run `docker compose down -v` on a real install, delete or `chown` a folder that already holds data without looking at it first and asking, or copy files into RustFS's data folder (it keeps its own on-disk format; move data through S3 only).
- **Ask before anything that costs money or exposes the service:** creating a Droplet or Space, opening a port, changing a router or DNS.
- **Don't open ports 8080, 9000, 9001 or 5341 to the internet.** The dashboard has no sign-in of its own.

## 1. Find out where it is going

Ask only what you can't see for yourself:

1. **Which machine?** A NAS, a Linux server, Docker Desktop, or a DigitalOcean Droplet. For anything but a plain Linux server, read `references/targets.md` now; it changes several steps below.
2. **Are you on that machine?** If not, run each command over `ssh <host>` and copy files with `scp` (`scp -O` on UGOS). Ask for the host name, never a password; SSH keys or their own prompt handle sign-in.
3. **Is DbnSync already there?** Look, don't just ask: run `docker compose ls` and `docker ps`, and look for a `dbnsync` folder with a `.env`. If it is there, read `references/existing-install.md` before changing anything: it covers updating it and trying a second copy beside it. For a repair, go to `references/troubleshooting.md`. The Compose file is named `dbnsync`, so a second copy in another folder takes over the running containers rather than sitting beside them.

Then check the machine and tell them what you found:

```sh
docker --version
docker compose version    # must be v2, with a space
df -h <the disk for the archive>
```

The defaults need about 90 GB on day one and grow about 750 GB a year, nearly all of it order books. If the disk is short, cut the roots or schemas rather than letting it fill. They need a Databento plan with history from a supported dataset; today that is CME Globex (`GLBX.MDP3`).

## 2. Get the deployment files

From a checkout of this repository, copy `Deploy/docker-compose.yml`, `Deploy/Caddyfile` and `Deploy/.env.example` to the target's folder, the last as `.env`, then `chmod 600 .env`. Otherwise:

```sh
mkdir -p ~/dbnsync && cd ~/dbnsync
curl -fsSLO https://raw.githubusercontent.com/squideyes/DbnSuite/main/Deploy/docker-compose.yml
curl -fsSLO https://raw.githubusercontent.com/squideyes/DbnSuite/main/Deploy/Caddyfile
curl -fsSL https://raw.githubusercontent.com/squideyes/DbnSuite/main/Deploy/.env.example -o .env
chmod 600 .env
```

Copy this skill's `scripts/` into a `scripts` folder beside them when you are working over SSH, so that the commands below work as written.

## 3. Make the archive folder

RustFS runs as user id 10001, so that user has to own the folder. Put it on the disk you checked.

```sh
sudo mkdir -p /srv/dbnsync/s3
sudo chown -R 10001:10001 /srv/dbnsync/s3
```

`sudo` asks for the password of the account they signed in with, and only they type it, so over SSH give them the whole command to run themselves: `ssh -t <host> "sudo chown -R 10001:10001 <folder>"`. It shows nothing as they type and prints nothing when it works, so check with `ls -ld <folder>` rather than asking them to run it again. If they don't know the password, stop there: it is theirs to find or reset in the NAS's own account settings. Don't work round it with a container running as root.

## 4. Fill in .env

Run `scripts/fill-secrets.sh .env`. It generates the S3 secret and Seq's password straight into the file, so neither passes through this conversation.

Set these yourself, since none is a secret:

- `S3_DATA_PATH`: the folder from step 3.
- `DBNSYNC_LOGS_URL`: Seq as their browser reaches it, such as `http://myserver:5341`.
- `DBNSYNC_ROOTS` and `DBNSYNC_SCHEMAS`, only if they want other than the defaults (twelve CME index, energy and metal roots; `mbo,ohlcv-1s,definition,statistics,status`).

Then have them put their Databento key in `.env`. Call it the settings file in the deployment folder, since the name means nothing to most people. The key is at https://databento.com/portal/keys and starts `db-`; it goes straight after `DATABENTO_API_KEY=`, with no spaces. Over SSH, give them the command that opens the file: `ssh -t <host> "nano <folder>/.env"`. In nano, Ctrl+U does not paste from the clipboard: they right-click in the terminal instead. Ctrl+O then Enter saves, and Ctrl+X closes.

Run `scripts/check-env.sh .env` and fix whatever it marks `FAIL`. Treat `PAID` as a stop: tell them spending is on and ask whether they meant that.

## 5. Start it

```sh
docker compose pull
docker compose up -d
```

If `DBNSYNC_IMAGE` names an image already loaded on this machine, there is nothing to pull: use `docker compose pull s3 seq` first.

## 6. Prove it works

Don't call it done until you have seen each of these, because a container that is merely up may still be refusing every job:

```sh
docker compose ps                        # all three running, s3 healthy
curl -s http://localhost:8080/health     # answers Healthy
curl -s http://localhost:8080/status     # the worker's state
docker compose logs --tail 50 dbnsync    # a pass has started; no refusals
```

These use the default port. If `DBNSYNC_PORT` is set, use that, or you will be checking some other install.

The first pass only submits jobs and stores nothing. Files come on the next pass, about five minutes after the first ends, so wait for `stored:` lines in the logs before calling it done.

Then report in plain words: the dashboard's address (`http://<server>:8080`), where the archive lives, and what happens next. With the defaults the first backfill is about 670 Databento jobs, at most 30 at a time, and takes a day or two; the dashboard's ETA gives the time to expect.

If anything fails, use `references/troubleshooting.md`.

## 7. Afterwards

Offer these, in this order, and do only what they accept. Skip them for a trial:

1. **Separate keys.** Until this is done the worker uses the S3 administrator's key, which can delete the archive. `references/keys.md`.
2. **Reach it from elsewhere.** `references/internet.md`. Recommend Tailscale unless they need to share with people outside their own devices.
3. **Set a Seq retention policy.** Seq keeps everything until told otherwise; 30 days is plenty, in its Settings.

To **update**, follow `references/existing-install.md`. Stopping loses nothing: the next pass picks up whatever was pending, and the archive is the only state.
Files in this package (8)
  • SKILL.md 9 KB
  • references/existing-install.md 5 KB
  • references/internet.md 2 KB
  • references/keys.md 2 KB
  • references/targets.md 4 KB
  • references/troubleshooting.md 3 KB
  • scripts/check-env.sh 4 KB
  • scripts/fill-secrets.sh 1 KB

Versions

The database has no other skills named dbnsync-deploy.

Database snapshot: 13.11.2026 · package version dd57f7270eba (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100pro-ssh
    Run commands on remote servers over SSH (AWS EC2, Synology NAS, any Linux host), by saved name or direct host/user/key. Use for 'check the server', 'look at logs', 'is the deploy live'.
  • 90digitalocean-containers-images
    DigitalOcean containers and images including Container Registry, preconfigured images, and custom images. Use when building, storing, or deploying container images or Droplet images on DigitalOcean.
  • 100synology-operations
    Use when the user wants to connect to their Synology NAS for file operations, share management, storage health checks, or general maintenance. Reads connection details from `$CLAUDE_USER_DATA/synology-mgmt/config.json` (populated by the `onboard` skill in this plugin). Triggers…
  • 100synology-nas
    Use this skill for ANY task on a Synology NAS via the mcp-server-synology MCP — managing files and folders on the NAS, searching NAS shares, reading or creating NAS files, running Download Station tasks (torrents, magnets, downloads), checking NAS health (disks, SMART, volumes,…
  • 100n8n-self-hosting
    Deploy a production self-hosted n8n end-to-end to a fresh Linux VM over SSH, using Docker Compose behind a Caddy reverse proxy with automatic HTTPS. Use whenever the user wants to self-host, install, set up, provision, or deploy n8n on their own server/VPS/box (Hetzner,…
  • 100n8n-self-hosting
    Deploy a production self-hosted n8n end-to-end to a fresh Linux VM over SSH, using Docker Compose behind a Caddy reverse proxy with automatic HTTPS. Use whenever the user wants to self-host, install, set up, provision, or deploy n8n on their own server/VPS/box (Hetzner,…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.