a3140babae
external_node script. |
||
---|---|---|
.. | ||
DESCR | ||
PLIST | ||
README |
$OpenBSD: README,v 1.3 2013/03/12 10:59:50 ajacoutot Exp $ +----------------------------------------------------------------------- | Running ${FULLPKGNAME} on OpenBSD +----------------------------------------------------------------------- The Puppet Dashboard has been installed into ${INSTDIR} Configuration is done in: ${INSTDIR}/config/settings.yml Database Configuration ====================== Before starting the Dashboard, a MySQL database must be configured in: ${INSTDIR}/config/database.yml The following commands can be used to create the database: # mysql -uroot -p<password> mysql> CREATE DATABASE dashboard_production CHARACTER SET utf8; mysql> CREATE USER 'dashboard'@'localhost' IDENTIFIED BY 'my_password'; mysql> GRANT ALL PRIVILEGES ON dashboard_production.* TO 'dashboard'@'localhost'; mysql> EXIT The MySQL maximum packet size needs to be increased to permit larger rows in the database; ${SYSCONFDIR}/my.cnf must be edited to include: max_allowed_packet = 32M Then the server must be restarted by running: # ${RCDIR}/mysqld restart Once the database has been setup and MySQL is up and running, the following command must be run to create the tables: # cd ${INSTDIR} && sudo -u _puppet-dashboard rake${MODRUBY_BINREV} RAILS_ENV=production db:migrate Note that this command *must* also be run after a Puppet Dashboard upgrade. Running the Dashboard ===================== These commands will start a Rails server on port 3000 then start the delayed job workers to asynchronously process resource-intensive tasks. They can be added to rc.conf.local(5) to start Puppet Dashboard at boot time. # cd ${INSTDIR} && sudo -u _puppet-dashboard ./script/server -d -e production 2>/dev/null # cd ${INSTDIR} && sudo -u _puppet-dashboard env RAILS_ENV=production script/delayed_job -p dashboard -n 4 -m start 2>/dev/null Puppet Configuration ==================== To use the Dashboard for reports, all agent nodes must submit reports to the master and the master must send reports to the Dashboard. ${SYSCONFDIR}/puppet/puppet.conf needs to be edited accordingly: [agent] report = true # this is the default [master] reports = store, http reporturl = http://dashboard.domain.tld:3000/reports/upload Node Classification ------------------- To use the Dashboard external node classifier (ENC) alongside Puppet DSL node definitions, the master should contain the following lines (to be adapted accordingly): [master] node_terminus = exec external_nodes = /usr/bin/env PUPPET_DASHBOARD_URL=http://localhost:3000 ${INSTDIR}/bin/external_node Maintenance =========== To import old Puppet reports that were created before the Dashboard got installed, the following command can be used: # cd ${INSTDIR} && sudo -u _puppet-dashboard rake${MODRUBY_BINREV} RAILS_ENV=production reports:import Cron Jobs --------- The MySQL database should be optimized on a montly basis. This can be done by adding the following entry to the root's crontab(5): 0 0 1 * * cd ${INSTDIR} && sudo -u _puppet-dashboard rake${MODRUBY_BINREV} RAILS_ENV=production db:raw:optimize Over time, old reports will start to accumulate which can slow down the Dashboard. Adding the following line to the root's crontab(5) will clean up reports older than 1 month and will run every night: 0 0 * * * cd ${INSTDIR} && sudo -u _puppet-dashboard rake${MODRUBY_BINREV} RAILS_ENV=production reports:prune upto=1 unit=mon Going further ============= Setting up followings is out of the scope of this README but should be considered when running Puppet Dashboard in production. - authentication (by default, anyone can access the Dashboard) - https (for encrypted communication) - Passenger (with Apache or Nginx for better performance) More information is available at: http://docs.puppetlabs.com/dashboard/manual/1.2/index.html