Menu
Enterprise RSS

Promtail agent

Promtail is an agent which ships the contents of local logs to a private Grafana Loki instance or Grafana Cloud. It is usually deployed to every machine that runs applications which need to be monitored.

Note

Promtail is feature complete. All future feature development will occur in Grafana Alloy.

It primarily:

  • Discovers targets
  • Attaches labels to log streams
  • Pushes them to the Loki instance.

Currently, Promtail can tail logs from two sources: local log files and the systemd journal (on ARM and AMD64 machines).

Log file discovery

Before Promtail can ship any data from log files to Loki, it needs to find out information about its environment. Specifically, this means discovering applications emitting log lines to files that need to be monitored.

Promtail borrows the same service discovery mechanism from Prometheus, although it currently only supports static and kubernetes service discovery. This limitation is due to the fact that Promtail is deployed as a daemon to every local machine and, as such, does not discover label from other machines. kubernetes service discovery fetches required labels from the Kubernetes API server while static usually covers all other use cases.

Just like Prometheus, promtail is configured using a scrape_configs stanza. relabel_configs allows for fine-grained control of what to ingest, what to drop, and the final metadata to attach to the log line. Refer to the docs for configuring Promtail for more details.

Support for compressed files

Promtail now has native support for ingesting compressed files. If a discovered target has decompression configured, Promtail will lazily decompress the compressed file and push the parsed data to Loki. The Promtail configuration example below shows how to to set up decompression:

yaml
server:
  http_listen_port: 9080
  grpc_listen_port: 0
positions:
  filename: /var/lib/promtail/positions.yaml
clients:
  - url: http://localhost:3100/loki/api/v1/push
scrape_configs:
- job_name: system
  decompression:
    enabled: true
    initial_delay: 10s
    format: gz
  static_configs:
  - targets:
      - localhost
    labels:
      job: varlogs
      __path__: /var/log/**.gz

Important details are:

  • It relies on the \n character to separate the data into different log lines.

  • The max expected log line is 2MB within the compressed file.

  • The data is decompressed in blocks of 4096 bytes. i.e: it first fetches a block of 4096 bytes from the compressed file and processes it. After processing this block and pushing the data to Loki, it fetches the following 4096 bytes, and so on.

  • It supports the following extensions:

    • .gz: Data will be decompressed with the native Gunzip Golang pkg (pkg/compress/gzip)
    • .z: Data will be decompressed with the native Zlib Golang pkg (pkg/compress/zlib)
    • .bz2: Data will be decompressed with the native Bzip2 Golang pkg (pkg/compress/bzip2)
    • .tar.gz: Data will be decompressed exactly as the .gz extension. However, because tar will add its metadata at the beginning of the compressed file, the first parsed line will contains metadata together with your log line. It is illustrated at ./clients/pkg/promtail/targets/file/decompresser_test.go.
  • .zip extension isn’t supported as of now because it doesn’t support some of the interfaces Promtail requires. We have plans to add support for it in the near future.

  • The decompression is quite CPU intensive and a lot of allocations are expected to occur, especially depending on the size of the file. You can expect the number of garbage collection runs and the CPU usage to skyrocket, but no memory leak is expected.

  • Positions are supported. That means that, if you interrupt Promtail after parsing and pushing (for example) 45% of your compressed file data, you can expect Promtail to resume work from the last scraped line and process the rest of the remaining 55%.

  • Since decompression and pushing can be very fast, depending on the size of your compressed file Loki will rate-limit your ingestion. In that case you might configure Promtail’s limits stage to slow the pace or increase ingestion limits on Loki.

  • Log rotations on compressed files are not supported as of now (log rotation is fully supported for normal files), mostly because it requires us modifying Promtail to rely on file inodes instead of file names. If you’d like to see support for it, create a new issue on Github asking for it and explaining your use case.

  • If you compress a file under a folder being scraped, Promtail might try to ingest your file before you finish compressing it. To avoid it, pick a initial_delay that is enough to avoid it.

  • If you would like to see support for a compression protocol that isn’t listed here, create a new issue on Github asking for it and explaining your use case.

Loki Push API

Promtail can also be configured to receive logs from another Promtail or any Loki client by exposing the Loki Push API with the loki_push_api scrape config.

There are a few instances where this might be helpful:

  • complex network infrastructures where many machines having egress is not desirable.
  • using the Docker Logging Driver and wanting to provide a complex pipeline or to extract metrics from logs.
  • serverless setups where many ephemeral log sources want to send to Loki, sending to a Promtail instance with use_incoming_timestamp == false can avoid out-of-order errors and avoid having to use high cardinality labels.

Receiving logs From Syslog

When the Syslog Target is being used, logs can be written with the syslog protocol to the configured port.

AWS

If you need to run Promtail on Amazon Web Services EC2 instances, you can use our detailed tutorial.

Labeling and parsing

During service discovery, metadata is determined (pod name, filename, etc.) that may be attached to the log line as a label for easier identification when querying logs in Loki. Through relabel_configs, discovered labels can be mutated into the desired form.

To allow more sophisticated filtering afterwards, Promtail allows to set labels not only from service discovery, but also based on the contents of each log line. The pipeline_stages can be used to add or update labels, correct the timestamp, or re-write log lines entirely. Refer to the documentation for pipelines for more details.

Shipping

Once Promtail has a set of targets (i.e., things to read from, like files) and all labels are set correctly, it will start tailing (continuously reading) the logs from targets. Once enough data is read into memory or after a configurable timeout, it is flushed as a single batch to Loki.

As Promtail reads data from sources (files and systemd journal, if configured), it will track the last offset it read in a positions file. By default, the positions file is stored at /var/log/positions.yaml. The positions file helps Promtail continue reading from where it left off in the case of the Promtail instance restarting.

API

Promtail features an embedded web server exposing a web console at / and the following API endpoints:

GET /ready

This endpoint returns 200 when Promtail is up and running, and there’s at least one working target.

GET /metrics

This endpoint returns Promtail metrics for Prometheus. Refer to Observing Grafana Loki for the list of exported metrics.

Promtail web server config

The web server exposed by Promtail can be configured in the Promtail .yaml config file:

yaml
server:
  http_listen_address: 127.0.0.1
  http_listen_port: 9080