Skip to content

Latest commit

 

History

History
138 lines (104 loc) · 4.31 KB

installation.md

File metadata and controls

138 lines (104 loc) · 4.31 KB

Installation

Requirements for building the project

  • Java 8+
  • Maven 3.5.x +
  • rpm-build on Mac OS X

Checkout the project

git clone https://github.com/apache/ambari-logsearch
cd ambari-logsearch

Build Log Search packages

  • a.) Create RPM packages (outputs: ambari-logsearch-assembly/target/rpms/*)
make rpm
  • b.) Create DEB packages (outputs: ambari-logsearch-assembly/target/*.deb)
make deb
  • c.) Create tarballs (outputs: ambari-logsearch-server/target/ambari-logsearch-portal.tar.gz and ambari-logsearch-logfeeder/target/ambari-logsearch-logfeeder.tar.gz)
make package

Install created packages

  • On RHEL / CentOS :
# Log Search Portal (server + UI)
yum install -y /my/path/ambari-logsearch-portal.rpm
# Log Feeder
yum install -y /my/path/ambari-logsearch-logfeeder.rpm
# or if a repo is available that contains logsearch packages
yum install -y ambari-logsearch-portal
yum install -y ambari-logsearch-logfeeder
  • On Debian / Ubuntu :
# Log Search Portal (server + UI)
dpkg -i /my/path/ambari-logsearch-portal.deb
# Log Feeder
dpkg -i /my/path/ambari-logsearch-logfeeder.deb
# or if a repo is available that contains logsearch packages
apt-get install ambari-logsearch-portal
apt-get install ambari-logsearch-logfeeder
  • With tarball :
mkdir logsearch
tar -xzf ambari-logsearch-portal.tar.gz -C logsearch

mkdir logfeeder
tar -xzf ambari-logsearch-logfeeder.tar.gz -C logfeeder

Install by Ambari UI

Click on Add Service page:

Add Service

Then select Log Search on Add Services page:

Add Service Wizard

Then (after setting the right configuration on Configure services page) you can start the Log Search / Log Feeder deployments.

Component names to use in Ambari Blueprints:

  • Log Search Portal: LOGSEARCH_SERVER (1 node, or 1+, as Log Search Portal is stateless)
  • Log Feeder: LOGSEARCH_LOGFEEDER(All nodes)

Requirements for using Log Search

  • Solr 6.x+ and ZooKeeper (as Solr is used in Cloud mode) is required for Log Search Portal - Ambari Infra Solr can be used as a built-in Solr service in Ambari, but of course it can use an external Solr as well.
  • Log Feeder in Cloud mode can work without Solr and ZooKeeper (that means logs will be shipped to HDFS/S3/GCS/WASB/ADLSv2 ... but Log Search Portal won't be able to display data from those Storages)

Configurations

Log Search Portal configuration:

Required configuration files for Log Search Portal at /usr/lib/ambari-logsearch-portal/conf:

  • logsearch.properties
  • logsearch-env.sh
  • log4j2.yml
  • info.properties
  • default.properties
Log Feeder configuration:

Required configuration files for Log Feeder at /usr/lib/ambari-logsearch-logfeeder/conf:

  • logfeeder.properties
  • logfeeder-env.sh
  • log4j2.yml
  • alias_config.json
  • global.config.json
  • output.config.json
  • input.config-<service_name>.json (use this filename pattern for input configurations)
  • checkponts/*.cp (checkpoint files - information about monitored input files)

See more about the configurations: Log Search properties, Log Feeder properties and Log Feeder shipper configurations

Starting Log Search / Log Feeders

Go to Ambari UI and start Log Search components or manually starts by logsearch.sh or logfeeder.sh script

# starting Log Search
/usr/lib/ambari-logsearch-portal/bin/logsearch.sh start
# starting Log Feeder
/usr/lib/ambari-logsearch-logfeeder/bin/logsearch.sh start