«

»

Jul 12 2014

NCC 1.1 is available for download Now and it boosts Nutanix Cluster Health

NCC is a framework that serves as the engine for the Nutanix Cluster Health. It consists of various modules and plugins. The modules are groups of plugins that correspond to a specific test category. NCC can be run as long as individual nodes are up, regardless of cluster state. The scripts run standard commands against the cluster or the nodes, depending on the type of information being retrieved.

 

Screen Shot 2014-07-12 at 9.10.12 PM

 

The philosophy of NCC is for each plugin to function as a test that can be interpreted in most cases as a clear pass/fail. In addition to [PASS] and [FAIL], there are two additional results. Warning [WARN] is used when a value returned is unexpected by the check and the value needs to be investigated. Informational [INFO] is used so that a plugin can return a value that cannot be evaluated as pass/fail and does not require any investigation.

NCC checks fall into two non-disjoint sets today: batch-run manual NCC checks, and NCC checks that drive Prism Health page (image below).  Not all the NCC tests are part of Prism 4.0 UI. There is a sub-set that are not good candidates to be part of cluster health for various reasons and can only be run by cli. The important thing to remember is that just downloading the new NCC release you are adding additional and improved health checks to you Nutanix cluster.

 

Screen Shot 2014-07-12 at 9.09.38 PM

 

(Click on the image to enlarge)

Screen Shot 2014-07-12 at 8.58.54 PM

 

Download NCC 1.1
http://download.nutanix.com/ncc/v1.1/nutanix-ncc-1.1.tar.gz

 

NCC Installation

  1. Download the tarball onto any CVM in the cluster.
    1. The directory to which you copy the tarball should exist on all nodes in the cluster (/home/nutanix is suggested).
    2. The directory should be owned by whatever accounts will use the NCC utility.
  2. Extract tarball:
    $ tar xvmf –recursive-unlink

    1. recursive-unlink option is needed to ensure old installs are completely removed.
  3. Execute the install script (provide the tarball name if it has been moved or renamed).
    $ ./ncc/bin/install.sh [-f ]

    1. Check the output of this command for any error messages. If it runs to success, you will see the following output

ncc_1-1_install

  1. Optionally, source the bash tab-completion script.
    $ source ~/ncc/ncc_completion.bash

    1. Currently completion only works for modules and plugins (does not support flags).
    2. To ensure tab completion works on subsequent logins, add the line to ~/.bashrc.

 

NCC’s release cycle is independent from the NOS release cycle. This enables Nutanix Engineering to develop new plugins and modules without creating a dependency on a particular NOS release. The release frequency is approximately one month, but it can vary depending on the amount of work that needs to be done to develop new plugins and modules.

You can find more information on NCC  at the Nutanix Portal (here)

 

This article was first published by Andre Leibovici (@andreleibovici) at myvirtualcloud.net.

7 comments

Skip to comment form

  1. comotoza

    What is Cluster Profile in NCC ?
    Power/Knowledge/Task/Other looks like VDI profiles.

  2. Andre Leibovici

    comotoza, VDI VMs are classified into following categories based on following common criteria:

    1)Task Worker: If vcpu’s < 2, memory_size < 2GB and disk < 20 GB 2)Knowledge Worker:If CPU's < 2, memory_size < 2GB and disk < 40 GB 3)Power Worker: If vcpu's < 3, memory_size < 4 GB and disk < 50 GB 4)Others This is used in scenarios where Nutanix is sold with the VDI Per Desktop Assurance. For more information go to http://www.nutanix.com/nutanix-vdi-assurance/

  3. comotoza

    I mean in NCC information about Cluster Profile but no any VDI inside.

    In Health we have sensor: Description: Verify whether the Controller VM is uplinked to the 10 GbE NIC.
    In scenario when only management and only vmotion vmkernels are in one network why Nutanix tried to use vmotion kernel? With NIOC I limit vmotion trafic and seen high Nutanix latency. Nutanix network profile have high share and no limit.

    Autopath tried to use not management vkernel but vmotion?

  4. Jim

    Andre,

    Do you know if there’s a script that can execute the NCC health_checks run_all command either as a bash script or powershell and export the result and then emailed.

    Thanks,
    Jim

  5. Andre Leibovici

    I have not seen any scripts for that yet. If you happen to create it let’s upload to the Nutanix Script Repo at https://github.com/nutanix.

    I will also let Steven Poitras, our scripting master, know about the need for that.

  6. stevenpoitras

    Already something in the works 🙂

  7. Andre Leibovici

    ‘The Dude’ to the rescue. Thanks for that.

Leave a Reply