Ambari-server / Ambari-agent automatic startups

to create new topics or reply. | New User Registration

This topic contains 1 reply, has 2 voices, and was last updated by  Larry Liu 2 years, 3 months ago.

  • Creator
    Topic
  • #15697

    Francois BORIE
    Participant

    Hi !

    Do you plan to set up automatic startup at boot for ambari-agents and ambari-server components ?

    It would be great if those components could have an entry in /etc/init.d/ directory and if we could be able to add them for automatic startups with chkconfig for example.

    Currently, if a host reboots for an unexpected reason, host is not reintegrated automatically to the HDP cluster.

    In addition, maybe the next step will be to also start Hadoop services that should be started by this ‘just-rebooted’ node.

    In our case, manual action is not the best choice. For example, if a node reboots during the night, it can cause more unavailability time than if the node reintegrated itself to the cluster. For a datanode, this is not a big problem. But for Namenode or Jobtracker daemons for example, it’s more critical (I don’t take the time to implement Namenode HA for the moment)

    Many thanks in advance for your answer and thanks for this great Hadoop distrib,

    Regards,

    Bobo

Viewing 1 replies (of 1 total)

You must be to reply to this topic. | Create Account

  • Author
    Replies
  • #15698

    Larry Liu
    Moderator

    Hi, Bobo

    Thanks for using HDP.

    Your idea is a great idea. Actually, we have this in our plan.

    At this moment as a workaround, you can try to put restart command in to /etc/rc.local if you are using centos/redhat.

    Here are steps:

    1. edit /etc/rc.local
    2. put the following lines
    ambari-server restart
    ambari-agent restart
    3. If you would like to start other hadoop services during reboot, please add startup commands to /etc/rc.local

    Hope this helps

    Thanks

    Larry

    Collapse
Viewing 1 replies (of 1 total)
Hortonworks Data Platform
The Hortonworks Data Platform is a 100% open source distribution of Apache Hadoop that is truly enterprise grade having been built, tested and hardened with enterprise rigor.
Get started with Sandbox
Hortonworks Sandbox is a self-contained virtual machine with Apache Hadoop pre-configured alongside a set of hands-on, step-by-step Hadoop tutorials.
Modern Data Architecture
Tackle the challenges of big data. Hadoop integrates with existing EDW, RDBMS and MPP systems to deliver lower cost, higher capacity infrastructure.