Installing Logagent on Docker

Certified Logagent Images¶

Please note the following instructions use the latest Logagent image from Docker Hub. If you want to use certified images please pull the image from Red Hat and Docker registries.

Red Hat Certified Image: docker pull registry.connect.redhat.com/sematext/logagent
Docker Certified Image: docker pull store/sematext/logagent

Installation for Docker¶

Logagent is a general purpose log shipper. The Logagent Docker image is pre-configured for the log collection on container platforms. It runs as a tiny container on every Docker host and collects logs for all cluster nodes and their containers. All container logs are enriched with Kubernetes and Docker EE/Swarm metadata.

See: sematext/logagent on Docker Hub.

Getting started¶

To run Logagent you will need a Logs App Token. If you don't have Logs Apps yet, you can create Apps now.

The Logagent docker container can be configured through the following environment variables:

REGION: Sematext Cloud region US or EU (default: US). The receiver URL will be set to EU/US default values. When using REGION, you don't need to set LOGS_RECEIVER_URL (see below).
LOGS_RECEIVER_URL: The URL of your Elasticsearch Endpoint (defaults to Sematext Cloud US https://logsene-receiver.sematext.com).
- For Sematext Europe use https://logsene-receiver.eu.sematext.com.
- For Elasticsearch https://elasticserch-server-name:9200.
LOGS_TOKEN: The index where the agent should log to (for Sematext Cloud users the logs token)
LOGAGENT_ARGS: Additional command line arguments for Logagent
```
LOGAGENT_ARGS="-n httpd"
```
to specify a log source name or
```
LOGAGENT_ARGS="-u 514"
```
to act as syslog server. Please refer to Logagent command line arguments in the Logagent Documentation
LOG_GLOB: Semicolon-separated list of file globs
```
/mylogs//*.log;/var/log//*.log
```
Mount your server log files into the container using a Docker volume e.g.
```
-v /var/log:/mylogs
```
-v /var/run/docker.sock:/var/run/docker.sock - Collect container logs by mounting the docker socket (mandatory)

Docker Run¶

The most basic start method is using docker run command:

docker pull sematext/logagent
docker run -d --restart=always --name logagent \
-e LOGS_TOKEN=YOUR_LOGS_TOKEN \
-e LOGS_RECEIVER_URL="https://logsene-receiver.sematext.com" \
-v /var/run/docker.sock:/var/run/docker.sock sematext/logagent

Docker Compose¶

To use Docker Compose create docker-compose.yml as follows and insert real tokens:

# docker-compose.yml
logagent:
  image: 'sematext/logagent:latest'
  environment:
    - LOGS_TOKEN=YOUR_LOGS_TOKEN
    - LOGS_RECEIVER_URL="https://logsene-receiver.sematext.com"
  cap_add:
    - SYS_ADMIN
  restart: always
  volumes:
    - '/var/run/docker.sock:/var/run/docker.sock'

Then start Logagent with the docker-compose file:

docker-compose up -d

Docker Swarm and Docker Enterprise¶

Connect your Docker client to Swarm or UCP remote API endpoint and deploy Logagent with following docker command with your Logs Tokens:

docker service create --mode global --name st-logagent \
--restart-condition any \
--mount type=bind,src=/var/run/docker.sock,dst=/var/run/docker.sock \
-e LOGS_TOKEN="REPLACE THIS WITH YOUR LOGS TOKEN" \
-e REGION=US \
sematext/logagent:latest

Kubernetes and OpenShift¶

Run Logagent as Kubernetes DaemonSet.

First, create logagent-daemonset.yml

curl -o logagent-daemonset.yml https://raw.githubusercontent.com/sematext/logagent-js/master/kubernetes/logagent-daemonset.yml

Then run the DaemonSet:

kubectl create -f logagent-daemonset.yml

On Red Hat OpenShift use the "oc" command instead of kubectl.

oc apply -f logagent-daemonset.yml

Kubernetes with containerd and IBM Cloud¶

Kubernetes can use cointainerd as container engine. In this case Logagent can't use the Docker remote API to retrieve logs and metadata. Instead logs are collected from containerd log files and requires access to the relevant directories. The logagent input-filter for containerd supports:

Tailing log files from /var/log/containers/, /var/log/pods and /var/data/kubeletlogs
Enrichment of logs with podName, namespace, containerName, containerId
Parsing containerd log headers (timestamp, stream, flags)
Parsing message content with logagent parser library

Run Logagent as Kubernetes DaemonSet.

First, create ibm-cloud-logagent-ds.yml

curl -o ibm-cloud-logagent-ds.yml  https://raw.githubusercontent.com/sematext/logagent-js/master/kubernetes/ibm-cloud-logagent-ds.yml

Set your Logs Token in the spec.env section in the ibm-cloud-logagent-ds.yml file.

Then run the DaemonSet:

kubectl create -f ibm-cloud-logagent-ds.yml

Mesos / Marathon¶

The following configuration will activate Logagent on every node in the Mesos cluster. Please note that you have to specify the number of Mesos nodes (instances) and Logs Token. An example call to the Marathon API:

curl -XPOST -H "Content-type: application/json" http://your_marathon_server:8080/v2/apps  -d '
{
  "container": {
    "type": "DOCKER",
    "docker": {
      "image": "sematext/logagent"
    },
    "volumes": [
      {
        "containerPath": "/var/run/docker.sock",
        "hostPath": "/var/run/docker.sock",
        "mode": "RW"
      }
    ],
    "network": "BRIDGE"
  },
  "env": {
        "LOGS_TOKEN": "YOUR_LOGS_TOKEN",
        "LOGS_RECEIVER_URL": "https://logsene-receiver.sematext.com"

  },
  "id": "sematext-logagent",
  "instances": 1,
  "cpus": 0.1,
  "mem": 100,
  "constraints": [
    [
      "hostname",
      "UNIQUE"
    ]
  ]
}

Configuration Parameters¶

Mandatory Parameters¶

Parameter / Environment variable	Description
LOGS_TOKEN	Logs Token enables logging to Sematext Cloud, see logging specific parameters for filter options and Log Routing section to route logs from different containers to separate Logs Apps (indices)
-v /var/run/docker.sock	Path to the docker socket

Optional Parameters¶

Parameter / Environment variable	Description
REGION	Sematext Cloud region US or EU (default: US). The receiver URL will be set to EU/US default values. When using REGION, you don't need to set LOGS_RECEIVER_URL (see below).
LOG_GLOB	Semicolon-separated list of file globs (e.g. /var/log/*/.log;/mylogs/*/.log) to collect log files from the host, assuming the log files are mounted to /mylogs using Docker -v /var/logs:/mylogs
LOGAGENT_ARGS	Additional command line arguments for Logagent (e.g. LOGAGENT_ARGS="-n httpd" to specify a log source name or LOGAGENT_ARGS="-u 514" to act as syslog server)
--privileged	The parameter might be helpful when Logagent could not start because of limited permission to connect and write to the Docker socket /var/run/docker.sock. The privileged mode is a potential security risk, we recommend to enable the appropriate security. Please read about Docker security: https://docs.docker.com/engine/security/security/
HTTPS_PROXY	URL for a proxy server (behind firewalls)
LOGS_RECEIVER_URL	URL for bulk inserts into Sematext Cloud. Required for Sematext Enterprise (local IP:PORT) or Sematext Cloud Europe: https://logsene-receiver.eu.sematext.com
JOURNALD_UPLOAD_PORT	Port number for the collection of journald logs, forwarded by systemd-journal-upload.service. Equals to Logagent argument --journald PORT.
SYSTEMD_UNIT_FILTER	A regular expression to filter journald logs by systemd unit name, e.g. "ssh.service\|docker.service". The default value is ".*".
CONFIG_FILE	Path to the configuration file, containing environment variables `key=value`. Default value: `/run/secrets/logagent`. Create a secret with `docker secret create logagent ./logagent.cfg`. Start Logagent with `docker service create --mode global --secret logagent --mount type=bind,src=/var/run/docker.sock,dst=/var/run/docker.sock sematext/logagent
--privileged	The parameter might be helpful when Logagent could not start because of limited permission to connect and write to the Docker socket /var/run/docker.sock. The privileged mode is a potential security risk, we recommend to enable the appropriate security. Please read about Docker security: https://docs.docker.com/engine/security/security/

Docker Logs Parameters¶

Parameter / Environment variable	Description
TAGGING_LABELS	A list of docker label names or environment variable names to tag container logs. Supporting wildcards. Default value: com.docker.,io.kubernetes.,annotation.io.*
IGNORE_LOGS_PATTERN	Filter logs by a JS regular expression. E.g. IGNORE_LOGS_PATTERN=\/healthcheck\|\/ping
LOGSENE_ENABLED_DEFAULT	Enables log collection for containers having no explicit label/environment variable LOGSENE_ENABLED set. Default value: true. See section Log Routing.

Whitelist Containers for Logging¶

Parameter / Environment variable	Description
MATCH_BY_NAME	Regular expression to white list container names
MATCH_BY_IMAGE	Regular expression to white list image names

Blacklist Containers¶

Parameter / Environment variable	Description
SKIP_BY_NAME	Regular expression to black list container names
SKIP_BY_IMAGE	Regular expression to black list image names for logging

Set Log Patterns¶

Logagent supports various log formats defined in patterns.yml file. The Log Parser Patterns can be customized by proving your YAML file.

Parameter / Environment variable	Description
PATTERNS_URL	Load pattern.yml via HTTP e.g. PATTERNS_URL=https://myserver/patterns.yml
LOGAGENT_PATTERNS	Pass patterns.yml via env. variable e.g. LOGAGENT_PATTERNS="$(cat ./patters.yml)"
LOGAGENT_PATTERNS_BASE64	Set to "true" if the LOGAGENT_PATTERNS patterns file you are passing in via env. variable is base64 encoded e.g LOGAGENT_PATTERNS_BASE64="$(cat ./patterns.yml \| base64)" . Useful if your patterns file is not getting set properly due to shell interpretation or otherwise.
PATTERN_MATCHING_ENABLED	Activate logagent-js parser, default value is true. To disable the log parser set the value to false. This could increase the throughput of log processing for nodes with a very high log volume.
-v /patterns.yml:/etc/logagent/patterns.yml	Mount a patterns file to customize patterns for log parsing

Other Options¶

Parameter / Environment variable	Description
-v /tmp:/log-buffer	Directory to store logs, in a case of a network or service outage. Docker Agent deletes these files after successful transmission.
GEOIP_ENABLED	Enables GeoIP lookups in the log parser, default value: "false"
GEOIP_FIELDS	A list of fields to perform geo IP lookups e.g. GEOIP_FIELDS="client_ip,remote_address
MAXMIND_LICENSE_KEY	Your MaxMind license key
MAXMIND_DB_DIR	Directory for the Geo-IP lite database, must end with /. Storing the DB in a volume could save downloads for updates after restarts. Using /tmp/ (ramdisk) could speed up Geo-IP lookups (requires add. ~30 MB main memory).
REMOVE_FIELDS	Removes fields from parsed/enriched logs. E.g. REMOVE_FIELDS=password,creditCardNo

Configuration Manual¶

Blacklisting and Whitelisting Logs¶

Not all logs might be of interest, so sooner or later you will have the need to blacklist some log types. This is one of the reasons why Logagent automatically adds the following tags to all logs:

Container ID
Container Name
Image Name
Docker Compose Project Name
Docker Compose Service Name
Docker Compose Container Number

Using this “log metadata” you can whitelist or blacklist log outputs by image or container names. The relevant environment variables are:

MATCH_BY_NAME — a regular expression to whitelist container names
MATCH_BY_IMAGE — a regular expression to whitelist image names
SKIP_BY_NAME — a regular expression to blacklist container names
SKIP_BY_IMAGE — a regular expression to blacklist image names

Some log messages are useless or noisy, like access to health check URLs in Kubernetes. You can filter such messages via regular expressions by setting the following environment variable:

IGNORE_LOGS_PATTERN=\/healthcheck|\/ping

Container Log Parsing¶

In Docker, logs are console output streams from containers. They might be a mix of plain text messages from start scripts and structured logs from applications. The problem is obvious – you can’t just take a stream of log events all mixed up and treat them like a blob. You need to be able to tell which log event belongs to what container, what app, parse it correctly in order to structure it so you can later derive more insight and operational intelligence from logs, etc.

Logagent analyzes the event format, parses out data, and turns logs into structured JSON. This is important because the value of logs increases when you structure them — you can then slice and dice them and gain a lot more insight about how your containers, servers, and applications operate.

Traditionally it was necessary to use log shippers like Logstash, Fluentd or Rsyslog to parse log messages. The problem is that such setups are typically deployed in a very static fashion and configured for each input source. That does not work well in the hyper-dynamic world of containers! We have seen people struggling with the Syslog drivers, parsers configurations, log routing, and more! With its integrated automatic format detection, Logagent eliminates this struggle — and the waste of resources — both computing and human time that goes into dealing with such things! This integration has a low footprint, doesn’t need retransmissions of logs to external services, and it detects log types for the most popular applications and generic JSON and line-oriented log formats out of the box!

Example: Apache Access Log fields generated by Logagent

For example, Logagent can parse logs from official images like:

Nginx, Apache, Redis, MongoDB, MySQL
Elasticsearch, Solr, Kafka, Zookeeper
Hadoop, HBase, Cassandra
Any JSON output with special support for Logstash or Bunyan format
Plain text messages with or without timestamps in various formats
Various Linux and Mac OSX system logs

Adding log parsing patterns¶

In addition, you can define your own patterns for any log format you need to be able to parse and structure. There are three options to pass individual log parser patterns:

Configuration file in a mounted volume: -v PATH_TO_YOUR_FILE:/etc/logagent/patterns.yml
Kubernetes ConfigMap example: Template for patterns.yml as ConfigMap
Content of the configuration file in an environment variable: -e LOGAGENT_PATTERNS=”$(cat patterns.yml)”
Set patterns URL as environment variable: -e PATTERNS_URL=http://yourserver/patterns.yml

The file format for the patterns.yml file is based on JS-YAML, in short:

– indicates an array element
!js/regexp – indicates a JavaScript regular expression
!!js/function > – indicates a JavaScript function

The file has the following properties:

patterns: list of patterns, each pattern starts with “-“
match: a list of pattern definition for a specific log source (image/container)
- sourceName: a regular expression matching the name of the log source. The source name is a combination of image name and container name.
- regex: JS regular expression
- fields: field list of extracted match groups from the regex
- type: type used in Sematext Cloud (Elasticsearch Mapping)
- dateFormat: format of the special fields ‘ts’, if the date format matches, a new field @timestamp is generated
- transform: A JavaScript function to manipulate the result of regex and date parsing

The following example shows pattern definitions for web server logs, which is one of the patterns available by default: Example from https://sematext.github.io/logagent-js/parser/

This example shows a few very interesting features:

Masking sensitive data with “autohash” property, listing fields to be replaced with a hash code
Automatic Geo-IP lookups including automatic updates for Maxmind Geo-IP lite database
Post-processing of parsed logs with JavaScript functions

The component for detecting and parsing log messages — logagent-js — is open source and contributions for even more log formats are welcome.

Log Routing¶

Routing logs from different containers to separate Sematext Cloud Logs Apps can be configured via Docker labels, or environment variables, e.g. on Kubernetes.

Tag a container with the label, or environment variable LOGS_TOKEN=YOUR_LOGS_TOKEN. Logagent inspects the containers for this label and ships the logs to the specified Logs App.

The following container environment variables and labels are supported:

LOGS_TOKEN=<YOUR_LOGS_TOKEN> - logs token for the container
LOGS_ENABLED=<true|false> - switch log collection for the container on or off. Note, the default value is configurable in Logagent configuration via the setting LOGSENE_ENABLED_DEFAULT.
LOGS_RECEIVER_URL=<URL, URL, URL> - set multiple log destinations. The URL should include the token or index of an Elasticsearch API endpoint. E.g. https://logsene-receiver.sematext.com/<YOUR_LOGS_TOKEN>

The following Kubernetes Pod annotations are equivalent:

sematext.com/logs-token=<YOUR_LOGS_TOKEN>
sematext.com/logs-enabled=<true|false>
sematext.com/logs-receiver-url=<URL, URL, URL>

Example: The following command will start Nginx webserver and logs for this container will be shipped to the related Logs App.

docker run --label LOGS_TOKEN=REPLACE_WITH_YOUR_LOGS_TOKEN -p 80:80 nginx

Or use environment variables for Kubernetes:

docker run -e LOGS_TOKEN=REPLACE_WITH_YOUR_LOG_TOKEN -p 80:80 nginx

All other container logs will be shipped to the Logs App specified in the docker run command for sematext/logagent with the environment variable LOGS_TOKEN.

By default, all logs from all containers are collected and sent to Sematext Cloud. You can change this default by setting the LOGSENE_ENABLED_DEFAULT=false label for the Logagent container. This default can be overridden, on each container, through the LOGS_ENABLED label.

Please refer to Docker Log Management & Enrichment for further details.

Known Issues¶

Conflict with Docker logging-drivers. Logagent is running with a valid Logs Token, but Sematext Cloud does not show container logs.

Please note that Logagent collects logs via Docker Remote API. If you use a Docker logging-driver other than the default json-file driver, logs will not be available via the Docker Remote API.

Please make sure that your container or docker daemon uses json-file logging driver. This ensures that logs are exposed via Docker Remote API. To check, run the "docker logs" command. If "docker logs CID" shows container logs then Logagent should be able to collect the logs as well.

Please check the parsed timestamps! Logs with timestamps in the future (or several months or years in the past) might not be displayed in Sematext Cloud.