The Hortonworks Community Connection is now live. A completely rebuilt Q&A forum, Knowledge Base, Code Hub and more, backed by the experts in the industry.

You will be redirected here in 10 seconds. If your are not redirected, click here to visit the new site.

The legacy Hortonworks Forum is now closed. You can view a read-only version of the former site by clicking here. The site will be taken offline on January 31,2016

Hortonworks Sandbox Forum

The 'namelog' service fails with err msg while starting in SANDBOX 1.3

  • #42558
    Rajeev Joshi

    I was trying out Hadoop (Sandbox 1.3 Version) for the first time, and was doing the Hello World Tutorial.

    Firstly, I got an error while installation because the clusterservices.txt file mentioned d: as log file repositories, and I had only a C: drive on my Server.
    I split the C: into E: (D: was the DVD Driver) with the sizes:
    C: 7.94GB free of 23.4 GB
    E: 15.6GB free of 16.4 GB
    Then, I modified the few lines at the top of clusterservices.txt to refer to E: instead of D:

    The installation was successful.

    When I reach the next step:


    One services failed, ‘namelog’.

    I looked at the logs but I could not figure out why the failure was happening.

    I am attaching the log directory for your help.
    2013-10-31 00:00:05,448 INFO org.apache.hadoop.hdfs.util.GSet: recommended=4194304, actual=4194304
    2013-10-31 00:00:05,462 INFO org.apache.hadoop.hdfs.server.common.Storage: Storage directory E:hdpdatahdfsnn does not exist
    2013-10-31 00:00:05,464 ERROR org.apache.hadoop.hdfs.server.namenode.FSNamesystem: FSNamesystem initialization failed.
    org.apache.hadoop.hdfs.server.common.InconsistentFSStateException: Directory E:hdpdatahdfsnn is in an inconsistent state: storage directory does not exist or is not accessible.
    at org.apache.hadoop.hdfs.server.namenode.FSImage.recoverTransitionRead
    THere is only a E:hdpdatahdfsdn directory in that directory.

    It will be most helpful if you can tell me what the error was, and where I went wrong.

    Awaiting an early reply,

    ~ rajeev joshi (e: (m: +91.98225.39873)

  • Author
  • #42569


    The Sandbox runs on CentOS, so I believe you must be installing HDP 1.3 for Windows – is that correct?
    You should probably place this into the windows forum where people would be best placed to help you.

    I would look through the installation logs to see why this directory wasn’t created correctly – or you can create the directory.



    Rajeev Joshi

    Hi Dave,
    1. How do I access the Windows forum?
    2. I tried creating the directory ‘nn’ (see the details above), but that did not solve the problem. I got something like “namenode is not Formatted…”

    I would really be VERY,VERY grateful if someone could help with the above problem, as I have spent literally hours just on the ‘Hello, World’ Windows tutorial without getting it correct, and that makes me feel very shameful and frustrated.
    Awaiting a reply,
    ~ rajeev (m:+91.98225.39873) (e:


    Hi Rajeev,

    You need to format the namenode after you create the directory – then the NameNode will start and services will connect successfully:

    hadoop namenode -format

    The windows forum is found here:



    Cheryle Custer

    If your goal is to just do the tutorial, please try the Sandbox rather than a full HDP installation. The tutorials are developed for the Sandbox and the Sandbox HUE interface. Find the Sandbox here: It’s a single node Hadoop environment that can run on your laptop.

    Rajeev Joshi

    Ms. Cheryle,
    I AM using the Sandbox as mentioned in the first line of the main mail above. The error is coming in the Sandbox code itself.

    In case the error is no reproducible to you, would you just note down the set of URLs I am supposed to access, and the choice on each page that I am to take, in solving the Hello WOrld Tutorial?

    Thanks a lot,

    Reply awaited,
    ~ rajeev (m:+91.98225.39873) (e:


    Hi Rajeev,

    From the description of your issue, you are not using the sandbox.
    The sandbox should be run in VirtualBox or VMWare and runs on CentOS 6
    The URL to access the sandbox is

    However, the namenode does not need formatting etc which is what you’ve had to do here.

    Please let me know the link you used to download the sandbox and how you are running it.



    Rajeev Joshi

    Hi Dave,
    Thanks for the prompt reply.
    The URL for getting the sandbox that I use is the FOURTH one mentioned in:
    On this web page, it is the Sandbox 1.3 for legacy testing that I have been using.
    The .ova file download successfully — albeit after 4 hours — and then I give it as input to the ESX client as an existing VM.
    The IP is not found, so the final screen does NOT say, “Access this through any browser.”
    S, I set the IP for the Sandbox VM by logging in as root/hadoop. And then the broswer part works correctly.

    Am I doing anything wrong or extra?

    Now, I load a W2k8R2 VM separately, and perform the installation tasks there.

    For this,, I go to the Hello World’ Tutorial, and download the HDP asper the ‘HortonWorks Data Platform Download’ hyperlink.
    THAT’s when I keep getting errors in the last-but-one step there

    Am I doing something wrong here? Pl let me know.

    Best Regards,
    ~ rajeev


    Hi Rajeev,

    Why are you installing a HDP for Windows cluster on a windows 2008 machine to use with ‘Sandbox’ ?
    The Sandbox is working and no-where in the ‘hello world’ tutorial does it ask you to install a Windows cluster.
    All you need to download is the sample data, then you can upload and manipulate it



    Rajeev Joshi

    Hi Dave,
    SO sorry, it was my mistake. I misread the last line of the first paragraph of ‘Hello World’ where it says that ‘ HDP is free to download and use in your enterprise and you can download it here: Hortonworks Data Platform Download’ and the last 4 words formed a hyperlink.

    I now know that no other machine needs to be involved, except the Sandbox.

    I understand that the screenshots for Apache Hive and Query Editor further down in the ‘Hello World’ tutorial refer to the Sandbox and the relevant .exe in the sandbox. Is that correct?

    I am away from my installation/laptop right now, so cannot continue further experimentation.

    I will be able to work further only on Wednesday 6thNov’13 as we are on a religious holiday here (equivalent in importance to ‘Christmas’ in the western world). Its called Diwali — the festival of Lamps.

    I thank you profusely for all your so very prompt answers, and will get back to you with hopefully a passed Smoke Test and a ‘Thumbs Up’ on the 6th.

    With Best Regards,
    ~ rajeev


    Hi Rajeev,

    No problem, the sandbox is totally standalone.
    If you wish to install HDP on another machine you must follow the install instructions and can format the namenode as I had asked earlier.
    This should clear up your issue.

    Let me know if you have any other problems.



The topic ‘The 'namelog' service fails with err msg while starting in SANDBOX 1.3’ is closed to new replies.

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.