Skip to content

docs: pipeline: inputs: systemd: document journal namespace option #2689

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
eschabell merged 1 commit into from
Sep 3, 2026
Merged

docs: pipeline: inputs: systemd: document journal namespace option #2689

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
53 changes: 50 additions & 3 deletions pipeline/inputs/systemd.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,18 +13,65 @@ The plugin supports the following configuration parameters:
| Key | Description | Default |
|:----------------------|:--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:--------|
| `db` | Specify the absolute path of a database file to keep track of the `journald` cursor. | _none_ |
| `db.sync` | Set a default synchronization (I/O) method. Possible values: `extra`, `full`, `normal`, and `off`. This flag affects how the internal SQLite engine synchronizes to disk. For more details, see the [SQLite documentation](https://www.sqlite.org/pragma.html#pragma_synchronous). Available in Fluent Bit v1.4.6 and later. | `full` |
| `db.sync` | Set a default synchronization (I/O) method. Possible values: `extra`, `full`, `normal`, and `off`. This flag affects how the internal SQLite engine synchronizes to disk. For more details, see the [SQLite documentation](https://www.sqlite.org/pragma.html#pragma_synchronous). Available in Fluent Bit v1.4.6 and later. | `full` |
| `lowercase` | Lowercase the `journald` field (key). | `false` |
| `max_entries` | When Fluent Bit starts, the Journal might have a high number of logs in the queue. To avoid delays and reduce memory usage, use this option to specify the maximum number of log entries that can be processed per round. Once the limit is reached, Fluent Bit will continue processing the remaining log entries once `journald` performs the notification. | `5000` |
| `max_fields` | Set a maximum number of fields (keys) allowed per record. | `8000` |
| `namespace` | Read from a specific `journald` namespace instead of the default journal. See [Journal namespaces](#journal-namespaces). Can't be used with `path`. Setting both causes a startup error. | _none_ |
Comment thread
eschabell marked this conversation as resolved.
| `path` | Optional path to the Systemd journal directory. If not set, the plugin uses default paths to read local-only logs. | _none_ |
| `read_from_tail` | Start reading new entries. Skip entries already stored in `journald`. | `false` |
| `strip_underscores` | Remove the leading underscore of the `journald` field (key). For example, the `journald` field `_PID` becomes the key `PID`. | `false` |
| `systemd_filter` | Perform a query over logs that contain specific `journald` key/value pairs. For example, `_SYSTEMD_UNIT=UNIT`. The `systemd_filter` option can be specified multiple times in the input section to apply multiple filters. | _none_ |
| `systemd_filter_type` | Define the filter type when `systemd_filter` is specified multiple times. Allowed values: `and`, `or`. With `and` a record is matched only when all `systemd_filter` have a match. With `or` a record is matched when any `systemd_filter` has a match. | `or` |
| `tag` | Fluent Bit uses tags to route messages. For the Systemd input plugin, tags have an additional capability: if the tag includes a wildcard (`*`), it will be expanded with the Systemd Unit file (`_SYSTEMD_UNIT`, like `host.* => host.UNIT_NAME`) or `unknown` (`host.unknown`) if `_SYSTEMD_UNIT` is missing. | _none_ |
| `systemd_filter_type` | Define the filter type when `systemd_filter` is specified multiple times. Allowed values: `and`, `or`. With `and` a record is matched only when all `systemd_filter` have a match. With `or` a record is matched when any `systemd_filter` has a match. | `or` |
| `tag` | Fluent Bit uses tags to route messages. For the Systemd input plugin, tags have an additional capability: if the tag includes a wildcard (`*`), it will be expanded with the Systemd Unit file (`_SYSTEMD_UNIT`, like `host.* => host.UNIT_NAME`) or `unknown` (`host.unknown`) if `_SYSTEMD_UNIT` is missing. | _none_ |
| `threaded` | Indicates whether to run this input in its own [thread](../../administration/multithreading.md#inputs). | `false` |

## Journal namespaces

Systemd can partition the journal into [namespaces](https://www.freedesktop.org/software/systemd/man/latest/systemd-journald.service.html#Journal%20Namespaces), where a separate `systemd-journald` instance keeps its own log storage for a set of units. Set `namespace` to collect from one of these namespaces instead of the default journal.

Keep the following in mind when using `namespace`:

- The plugin reads only the named namespace. Entries in the default journal aren't collected. To collect both, define a separate `systemd` input for each.
- `namespace` and `path` are mutually exclusive. If you set both, Fluent Bit logs `path and namespace are mutually exclusive` and fails to start.
- This option requires a Fluent Bit binary built against `libsystemd` 245 or later, which is when `sd_journal_open_namespace` was introduced. If the binary was built against an older version, Fluent Bit logs `namespace requires libsystemd >= 245` and fails to start. Use `path` instead on those builds.
- When you use `db` to track the journal cursor, give each namespace its own database file. Reusing one database file across namespaces produces a cursor that the other namespace can't seek to, and the plugin logs a `seek_cursor failed` warning and starts over.

The following example collects logs from a namespace named `logging`, and writes the cursor to a database file dedicated to that namespace:

{% tabs %}
{% tab title="fluent-bit.yaml" %}

```yaml
pipeline:
inputs:
- name: systemd
tag: host.*
namespace: logging
db: /var/lib/fluent-bit/systemd-logging.db
outputs:
- name: stdout
match: '*'
```

{% endtab %}
{% tab title="fluent-bit.conf" %}

```text
[INPUT]
Name systemd
Tag host.*
Namespace logging
DB /var/lib/fluent-bit/systemd-logging.db

[OUTPUT]
Name stdout
Match *
```

{% endtab %}
{% endtabs %}

## Get started

To receive Systemd messages, you can run the plugin from the command line or through the configuration file.
Expand Down
Loading