Nutanix: Clean up system partition /home/nutanix

I have seen many AOS update pre-checks failed, simply because there was not enough space left on the system partition.
This looks like this (“Not enough space on /home/nutanix partition”):

Meanwhile there is also a regular check within PRISM, which will give you a warning if a CVM reach more as 75% utilization:

This happens usually due to left over install binaries from previous version or from large log bundles. The task to free up space you need to accomplish mostly via CLI. The alarm recommends you to delete unnecessary files via the AOS upgrade menu, but this only works with binaries which are newer as the current installed version. In all my cases the old ones did eat up the space, which are no longer shown in PRISM, so let´s turn over to the CLI.

With a simple df command in combination with allssh you can do a quick check on the actually utilization of /home/nutanix and as you can see its quite easy to figure out which of the two CVMs reached the 75% (also includes in the PRISM alarm details):

Run the following command on any CVM in the cluster to get an list of the 10 largest files in the /home/nutanix directory on all CVMs.

allssh 'sudo du -aSxh /home/nutanix/ | sort -h | tail -n 10'

As you can see in the following example output I found NOS/AOS binaries from version 4.7.x and 5.0.x on two of the CVMs and one large NCC output:

So just go ahead and delete those files to free up the space. Please note that you should keep the current AOS binaries you have installed on your cluster (used by foundation eg. if you expand your cluster etc.)

NOTE: Please use the commands above only if you know what you are doing and at your own risk. If you are uncertain, I would strongly recommend involving the Nutanix support.

Speak Your Mind

*

This site uses Akismet to reduce spam. Learn how your comment data is processed.