#netapp_info , get aggregates sort by freespace

1 messages · Page 1 of 1 (latest)

icy hatch
#

hello, anyone already dealed with netapp_info try to get the aggregate with the most free space available
? i'm playing arround with netapp_info but with no success so far

clear kiln
#

Not grabbing the one with the most free space but this is what I am using:

      netapp.ontap.na_ontap_rest_info:
        gather_subset:
          - storage/aggregates
        parameters:
          snaplock_type: non_snaplock
          state: online
          #block_storage.storage_type: ssd  # Added in ONTAP 9.11
          space.block_storage.physical_used_percent: "<85"
        fields:
          - 'space.block_storage.physical_used_percent'
      register: aggr_info

    - name: Debug - Aggregates found
      debug:
        msg: "{{ aggr_info }}"
        verbosity: 2

    - name: Creating list of eligible aggregates
      set_fact:
        aggr_list: "{{ aggr_list + [ item.name ] }}"
      with_items: "{{ aggr_info.ontap_info['storage/aggregates'].records }}"

    - name: Debug - Aggregate list
      debug:
        msg: "{{ aggr_list }}"
        verbosity: 1

    - name: Select random aggregate from list
      set_fact:
        aggr_name: "{{ item }}"
      with_random_choice: "{{ aggr_list }}"```

Then using "aggr_name" in the "volume create" task, this way it fails if there are no eligible aggregates.
Would love to see a nicer way of doing it though.
heavy moat
#

I have a role, that does that, here is what those tasks would look like in a playbook. Just get the JSON from ontap_rest_info, extract the values you need and sort the new list by the available field

# 
- name: 'get least full aggregate'
  hosts: all
  connection: local
  gather_facts: false
  collections:
    - netapp.ontap
  vars:
    netapp_hostname: "{{ inventory_hostname }}"

  tasks:
    - name: 'get ontap info'
      na_ontap_rest_info:
        hostname: "{{ netapp_hostname }}"
        username: "{{ netapp_username }}"
        password: "{{ netapp_password }}"
        https: true
        validate_certs: false
        use_python_keys: true
        gather_subset:
          - storage/aggregates
        fields:
          - 'name'
          - 'node.name'
          - 'space.block_storage.size'
          - 'space.block_storage.used'
          - 'space.block_storage.available'
      register: aggregate_info

    - name: 'create clean list of dicts from aggregate json'
      set_fact: 
        aggregate_info: "{{ aggregate_info| json_query(jmespath_expression) }}"
      vars:
        jmespath_expression: "ontap_info.storage_aggregates.records[*].{ name: name, space_total: space.block_storage.size, space_used: space.block_storage.used, space_available: space.block_storage.available }"
        
    - name: 'sort the list by space_available'
      set_fact:
        aggregate_info: "{{ aggregate_info | sort(attribute='space_available', reverse=true) }}"   

    - name: 'write the dict with the most free space to least_full_aggregate'
      set_fact:
        least_full_aggregate: aggregate_info[0]
#

you can also do maths afterwards to check wether the volume you want to provision would still fit on that aggregate and at what percentage the aggregate would be afterwards. That's why I included the size, used fields in the dict

icy hatch
#

thank you @clear kiln and @heavy moat for feedback!
awesome work - @heavy moat your playbook fits perfect for my use case and it is working fine. thank you!

worldly fjord
#

@clear kiln @heavy moat @icy hatch
I got errors on "aggr_list" variable is not defined. Can you please share the line on how to define “aggr_list” variable in @clear kiln playbook? Thank you!

worldly fjord
#

Anybody please help me out on the question above?

primal path
#

From this playbook #1103301527650717866 message

It looks like the first play is to rest info and it registers aggr_info
And the second play uses that aggr_info.

I would have to see what your playbook looks like to see what causing the issue for you

worldly fjord
#

@primal path Thanks for your response.

The error is not on "aggr_info", but "aggr_list", as shown below, the variable was not defined before used, and caused errors. My playbook is the same. Should we define it before use? How and in where?
- name: Creating list of eligible aggregates set_fact: aggr_list: "{{ aggr_list + [ item.name ] }}" with_items: "{{ aggr_info.ontap_info['storage/aggregates'].records }}"

TASK [Creating list of eligible aggregates] *************************************************************************************************************************************************************************************************
fatal: [localhost]: FAILED! => {"msg": "The task includes an option with an undefined variable. The error was: 'aggr_list' is undefined\n\nThe error appears to be in './aggr_freespace1.yml': line 32, column 7, …

unique thorn
worldly fjord
#

@unique thorn It works, Thanks!
Just checked the link, and understand how "{{ some_variable | default(5) }}" works, in general.

But, what I don't understand about is that what variable "[]" inside "default([])" was actually defined? and I don't see a value as default in it. ?

clear kiln
#

Sorry, forgot to mention that I define "aggr_list" as empty at the beginning of the playbook:

    aggr_list: []```
But Mohans solution looks much better. 😃
worldly fjord
#

aggr_list: [] is a list type or dict type?

clear kiln
round cargo