fatal: HTTP Error 500 warning cause playbook to fail

Question

fatal: HTTP Error 500 warning cause playbook to fail

freedge opened this issue 4 years ago · 7 comments

SUMMARY

Running a playbook for volume creation, got a fatal error for something that is actually, a warning.

fatal: [blade1]: FAILED! => {"changed": false, "msg": {"code": 500, "data": {"easytier": "on", "iogrp": "io_grp0", "mdiskgrp": "IaaS_1", "name": "blade1_boot_0", "rsize": "2%", "size": "200", "unit": "gb"}, "err": ["HTTPError %s", "HTTP Error 500: Internal Server Error"], "out": "\"Command error occurred; error code 1: CMMVC6372W The virtualized storage capacity that the cluster is using is approaching the virtualized storage capacity that is licensed. \"", "url": "https://svc:7443/rest/mkvdisk/-autoexpand"}}

The playbook failed because of this fatal error, however the disk was well created! After a rerun, playbook completed successfully.

I expect the API (Current software level: Version 8.2.1.8 (build 147.19.1911061502000) )
to not return an error 500 in case of warning.

If this cannot be fixed easily, I would expect at least the ansible module to fail when running in check mode (here, nothing was reported)

ISSUE TYPE

Bug Report

COMPONENT NAME

ibm_svc_vdisk

ANSIBLE VERSION

ansible 2.10.0

Answer 1 · 2020-09-18T05:29:47.000Z

Can you please share the playbook ?

Answer 2 · 2020-09-18T06:59:19.000Z

it should work with just the example playbook for ibm_svc_vdisk

Set the SCU low enough and run something like this

---
- name: Create volume
  hosts: blade
  connection: local
  gather_facts: no
  serial: 1
  collections:
    - ibm.spectrum_virtualize
  vars:
    svcuser:       "{{ lookup('env','IBMSVC_USERNAME') }}"
    svcpassword:   "{{ lookup('env','IBMSVC_PASSWORD') }}"
    volume_name:   "{{ inventory_hostname}}_boot_2"
  tasks:
    - name: Create volume
      ignore_errors: false
      delegate_to: localhost
      ibm_svc_vdisk:
        clustername: '{{ svc_ip }}'
        username: "{{ svcuser }}"
        password: "{{ svcpassword }}"
        log_path: /tmp/playbook.debug
        name: "{{ volume_name }}"
        state: present
        mdiskgrp: '{{ svc_pool }}'
        size: "200"
        unit: gb

It fails with error

fatal: [blade1]: FAILED! => {"changed": false, "msg": {"code": 500, "data": {"easytier": "off", "mdiskgrp": "pool", "name": "blade1_boot_2", "size": "200", "unit": "gb"}, "err": ["HTTPError %s", "HTTP Error 500: Internal Server Error"], "out": "\"Command error occurred; error code 1: CMMVC6372W The virtualized storage capacity that the cluster is using is approaching the virtualized storage capacity that is licensed. \"", "url": "https://svc:7443/rest/mkvdisk"}}

Answer 3 · 2020-09-18T10:51:13.000Z

Thanks, I asked for the playbook because I spotted the "rsize" parameter in the output log in your first comment.
Because our module (ibm_svc_vdisk) does not support "rsize" yet.

Nevertheless, We will analyse and look into the issue.

Answer 4 · 2020-09-18T12:08:01.000Z

haha yes, I had to fork and make a few changes. I can submit pr if you decide to accept contributions...
on vdisk I need the auto expand and iogroup options, on vol_map the scsi option to specify the lun id.

Answer 5 · 2020-09-18T15:29:42.000Z

Thanks man, will let you know once we start accepting contributions. :)

Answer 6 · 2021-12-29T05:35:05.000Z

The issue is not reproducible on the latest ibm_svc_vdisk and ibm_svc_manage_volume module. The Ansible module throws proper error message in such case and volume is also not created.

Answer 7 · 2022-01-12T07:32:16.000Z

Since the issue is not reproducible anymore on the latest code hence closing this ticket.