Skip to content

Instantly share code, notes, and snippets.

@xterat

xterat/Run-Crail-on-Ubuntu-16.04.md

Last active February 5, 2018 02:47
Show Gist options
  • Save xterat/b9bb991fd0016d897db731da65aefa48 to your computer and use it in GitHub Desktop.
Save xterat/b9bb991fd0016d897db731da65aefa48 to your computer and use it in GitHub Desktop.
Download ZIP
Run Crail on Ubuntu 16.04
Raw
Run-Crail-on-Ubuntu-16.04.md

Run crail on Ubuntu 16.04

  1. Set HugePages_Total by:

    sudo bash -c "echo 81920 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages"

    Verify it by cat /proc/meminfo.

  2. Mount hugetlbfs:

    sudo mount -t tmpfs -o size=10g tmpfs /media/crail

  3. Set up index path's mount point.

    sudo mount -t hugetlbfs none /mnt

  4. Set up CRAIL_HOME.

  5. Change mount point's owner:

    sudo chown -R hadoop:hadoop /mnt
sudo chown -R hadoop:hadoop /media/crail

  6. Run crail's namenode & datanode.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment