HDP on Linux – Installation Forum

upgrade to 1.3

  • #27036
    Andrew Hume

    in the upgrade instructions, step 2 says (on RHEL6) to do an
    rpm -e –npostun oozie $old_version_number

    what the heck are we suppose dto use for $old_version_number?
    it clearly isn’t any version number, for the rpm command is expecting a package name.

    also, i hate instructions that just say “do X, and then wait until some status has change dto Y”.
    you really need to give an estimate of how long this should take, or how to verify that some progress is actually being made. for example, i’ve been waiting for over an hour and can’t find any log file that is being updated. is something broken? (rhetorical!) how would or could i know?

to create new topics or reply. | New User Registration

  • Author
  • #27114

    Hi Andrew,

    Firstly let me apologize for not getting back to you sooner on this. You are correct that rpm is expecting a package name. You can find the package name with “rpm -qa | grep oozie”. The instructions should have pointed this out. There is also a typo in what is there, a ‘-‘ is missing between oozie and $old_version_number. Thanks for bringing these to our attention, I’ll log a bug about them so that they will be fixed. As far as the estimation on the time goes though it is quite difficult to come up with one, as there are a large number of variables that can cause it to vary greatly. Such as the number of computer in the cluster, the speed of the network, the speed of the computers and many more.


    Andrew Hume


    i would point out that the actual answer was that despite following the instructions, the software wasn’t actually updated and so the hdfs release never changed.

    but in answer to your point about times being variable, i would point out that the time for modern hardware is of teh order of seconds to minutes per TB, and that in fact there is a specific log message to look out for. this could be mentioned in the instructions.

    its all fixed now, and hopefully the instructions will continue to improve.



    Hi Andrew,

    Thanks for letting us know that “it’s all fixed” now. Sorry that you had the trouble. I agree with your hope that the documentation will continue to improve.


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

Support from the Experts

A HDP Support Subscription connects you experts with deep experience running Apache Hadoop in production, at-scale on the most demanding workloads.

Enterprise Support »

Become HDP Certified

Real world training designed by the core architects of Hadoop. Scenario-based training courses are available in-classroom or online from anywhere in the world

Training »

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.