Question

Storage summary

  • 15 February 2019
  • 6 replies
  • 4932 views

Userlevel 1
Badge +5
Hi Nutanix Community,
We often wonder what is the real storage capacity available on a Nutanix cluster.

We have a three node cluster, in RF2.
In Prism Element main page, under "Storage Summary", I see : 20.87TiB free (logical) of 22.37 TiB
https://
Data Resiliency is OK. (green)

Which one of the following statements is correct ?
  • I can fill my storage up to 22.37 TiB without losing the data resiliency (=I can lose a node without losing data).
  • I can fill my storage only up to 66% of 22.37Tib (= 14.76Tib) without losing the data resiliency
Thank you 🙂

This topic has been closed for comments

6 replies

Userlevel 6
Badge +5

Hi Infomil,

The logical space available takes into account the replication factor (RF2 or RF3) that is configured on the cluster which defines the number of copies of the data to be stored. For RF2 that would be two copies and for RF3 three copies of data.

On a cluster level for 20.87TiB free (logical) out of 22.37 TiB you can use 20.87TiB. It is not recommended to go above 85-90% utilisation.

For a storage container note that if space reservation is not enabled (space is not guaranteed for the container) then if the free space of the storage pool is less then the same of the container, free space on the container will be equal to the free space of the storage pool. Something to keep in mind when working with multiple storage containers on the same pool.

Badge

Dear All, 

First of all, thank you for opening this thread. 

Is there an option or setting to tell the storage summary diagram that please draw a line like: ‘Dear Operator, with the current storage settings you can utilize your cluster till XY TB/GB/whatever?’ I think many Nutanix Cluster will be operated by engineers and technicians with different level of Nutanix knowledge and I have a fear that not everybody can take care of this. 

Thanks,
rchrd

Hi,

 

I have heard that it will be included in a future version. I was told this by a trainer in a course. Have unfortunately no KB article or similar.

 

Best regards

Erik

Dear All, 

First of all, thank you for opening this thread. 

Is there an option or setting to tell the storage summary diagram that please draw a line like: ‘Dear Operator, with the current storage settings you can utilize your cluster till XY TB/GB/whatever?’ I think many Nutanix Cluster will be operated by engineers and technicians with different level of Nutanix knowledge and I have a fear that not everybody can take care of this. 

Thanks,
rchrd

Userlevel 1
Badge +1
Hi All,

In the last 4 months, I experienced two similar cases with different customers who used up too much of their available storage causing their cluster to show Critical status for Data Resiliency.

Please check out the following KB article regarding the recommended guidelines for maximum storage utilization on a Nutanix cluster: https://portal.nutanix.com/#/page/kbs/details?targetId=kA0600000008ducCAA.

That article provided me with the required formula to calculate the exact threshold on storage utilization beyond which a Nutanix cluster becomes Not Data Resilient anymore.

I hope this helps you out as well.

Best Regards,
Vishwapreet
Userlevel 1
Badge +5
We have tested that.
The conclusion, in my scenario is: when I go below 36.5% of free storage, the Data Resiliency Status goes from OK to Critical.
With the following warnings:
"Cluster can not tolerate node failure and guarantee available rebuild capacity"
"Metadata ring partitions with nodes: cvmIP1, cvmIP2, cvmIP3 are not fault tolerant."
Badge
Our experience is that you will only be able to fill storage up to a percentage of the advertised free space (statement #2). As far as the exact percentage, or how to calculate it ... I'm not sure. Is it Total Storage All Nodes - 1? Is there overhead that needs to be accounted for? That's kind of what led me here, looking for those answers.

Thin vs thick provisioning, physical vs logical, actual data used, runway ... I'm still trying to find meaningful and accurate reports that will help us monitor usage and alert us before we hit that tipping point.