Duplicati Dashboard is a monitoring solution for Duplicati.
It allows you to monitor your backups status, collects stats for each reports and alerts you by email when a backup fails. It is intended to be self-hosted and works with docker-compose.
Everything is already pre-configured and ready to be deployed.
The right side of the video is not integrated in the dashboard. You can't control your backup with it.
git clone https://github.com/fabien-github/duplicati_dashboard.git
cd duplicati_dashboard
docker-compose up -dAdd theses two options for each backup you want to monitor:
- send-http-result-output-format: json
- send-http-url: http://localhost:8080
This assumes that your Duplicati instance is on the same host as your Duplicati Dashboard.
Login: admin Password: password
For email alerting, you need to configure a SMTP relay. See Configuration > Env file
The file config.env is used to configure some options. It will be shared between the 3 containers.
Only use this default configuration for testing purposes.
Influxdb variable will create and setup the database only on the first startup.
| Variables | Default | Description |
|---|---|---|
| DOCKER_INFLUXDB_INIT_MODE | setup | Automatically bootstrap the system |
| DOCKER_INFLUXDB_INIT_USERNAME | telegraf_user | Influxdb superadmin user |
| DOCKER_INFLUXDB_INIT_PASSWORD | telegraf_password | Influxdb superadmin password |
| DOCKER_INFLUXDB_INIT_ORG | telegraf_org | Influxdb Organization (used by influxdb / telegraf / grafana) |
| DOCKER_INFLUXDB_INIT_BUCKET | telegraf | Influxdb bucket to store reports (used by influxdb / telegraf / grafana) |
| DOCKER_INFLUXDB_INIT_ADMIN_TOKEN | telegraf_token | Influxdb superadmin token (used by influxdb / telegraf / grafana) |
| DOCKER_INFLUXDB_INIT_RETENTION | Influxdb data retention, default will retain forever | |
| INFLUXD_REPORTING_DISABLED | false | Disable InfluxData telemetry |
| TELEGRAF_LISTENER_PORT | 8080 | Port used by http_listener_v2 input, endpoint for the reports sent by Duplicati |
| TELEGRAF_LISTENER_PATH | / | Path to listen to |
| GF_SECURITY_ADMIN_USER | admin | Grafana superadmin user |
| GF_SECURITY_ADMIN_PASSWORD | password | Grafana superadmin password |
| GF_SERVER_ROOT_URL | http://localhost:3000 | Grafana URL, used in some templates like email notifications |
| GF_DASHBOARDS_DEFAULT_HOME_DASHBOARD_PATH | /etc/grafana/provisioning/dashboards/duplicati_dashboard.json | Force Duplicati dashboard by default on home page |
| GF_SMTP_ENABLED | false | Set to true for email notifications |
| GF_SMTP_HOST | localhost:25 | SMTP relay server. [host]:[port] |
| GF_SMTP_FROM_NAME | Grafana | Name of the email sender |
| GF_SMTP_USER | In case of SMTP auth | |
| GF_SMTP_PASSWORD | In case of SMTP auth | |
| GF_SMTP_FROM_ADDRESS | admin@grafana.localhost | Address used when sending out emails |
| GF_SMTP_EHLO_IDENTITY | ${HOSTNAME} | Name to be used as client identity for EHLO in SMTP dialog (Default will be the container ID) |
| GF_SMTP_STARTTLS_POLICY | “OpportunisticStartTLS”, “MandatoryStartTLS”, “NoStartTLS” | |
| NOTIFIER_EMAIL_RECIPIENT | example@example.com | Recipients for email notification (separated by a semicolon) |
| NOTIFIER_EMAIL_REMINDER_ENABLE | true | Re-send an email if alerts are still active |
| NOTIFIER_EMAIL_REMINDER_FREQUENCY | 2h | Delay between email reminders |
The dashboard is locked by the Grafana provisioning system. You can't edit the datasource, the dashboard or the alert notifier from the UI. You will need to copy the dashboard or disable the provisioning configuration.
Dashboard path: ./grafana/provisioning/dashboards/duplicati_dashboard.json
The idea is to keep the stack easy to deploy for everyone without investing time to learn Grafana configuration.
Feel free to fork the project or directly edit files on your own.
More information here.
- Grafana will discover your backups name from the reports but only over the last 90 days. So if your backups are scheduled for more than 90 days, you will need to edit the request of the variable
Backupin the dashboard configuration:
from(bucket: v.defaultBucket)
|> range(start: -90d)
|> filter(fn: (r) => true)
|> toString()
|> group(columns: ["backup-name"])
|> distinct(column: "backup-name")
|> keep(columns: ["_value"])- Last reported status / Last reported variations / Alerting graph are based over the past 30 days. You will need to adapt each panel requests if your backups are scheduled over more than 30 days.
import "influxdata/influxdb/schema"
from(bucket: v.defaultBucket)
|> range(start: -30d)
...The section "Alerting graph" is only used to trigger an alert when a backup fails. This is due to the lack of grafana alert support on other panel type. #6983
Backups status will be checked every minutes. An alert will be triggered after a pending status of 10min. Same delays are used on the recovery.
Warning reports don't trigger an alert.
After getting inside the influxdb container:
influx delete \
--org telegraf_org \
--bucket telegraf \
--token "telegraf_token" \
--start 1970-01-01T00:00:00Z \
--stop $(date +"%Y-%m-%dT%H:%M:%SZ") \
--predicate '_measurement="duplicati" AND "backup-name"="backup_to_delete"'-
Telegraf: Receive JSON reports from Duplicati.
-
Influxdb: Store reports converted by Telegraf.
-
Grafana: Requests Influxdb to generate dashboard and alerts.
Telegraf endpoint provides a limited HTTP autentification.
The configuration file is located here : ./telegraf/telegraf.conf
Feel free to add a proxy like traefik or nginx to protect the stack on an unsecure network. (TLS, IP Restrictions, ...)
- Not sure of the scalabilty, requests to the database are not efficients. Timeseries databases are not really adapted for this kind of data. This is mainly due to the nested and uneven json format from the reports and the variation time between reports.
- Features are limited directly by the stack itself. For example, it's nearly impossible to add a managment system for the backups.
- This project has no link with the development of Duplicati and his team.
Distributed under the GNU General Public License v3.0 License. See LICENSE for more information.