Shelf FW download issues | NetApp | Page 1

eternal field Jan 25, 2024, 8:29 AM

#

Hi guys,
I'm using a playbook to distribute teh shelf fw to all of our clusters.
I see in Ansible the package is downloaded and also in the audit.log of teh cluster I can see the download. But afterwards, the new fw is not available for updates neither it is started automatically.

task:

  netapp.ontap.na_ontap_firmware_upgrade:
    hostname: "{{ netapp_cluster }}"
    username: "<user_name>"
    password: "{{ pwd_definition.ansible_usr }}"
    https: true
    validate_certs: false
    package_url: "<download url of all_shelf_fw.zip>"
    force_disruptive_update: false
    firmware_type: shelf
    use_rest: always
  ignore_errors: true```

#

output of -vvv

    "changed": true,
    "invocation": {
        "module_args": {
            "cert_filepath": null,
            "clear_logs": true,
            "disk_fw": null,
            "fail_on_502_error": false,
            "feature_flags": null,
            "firmware_type": "shelf",
            "force_disruptive_update": false,
            "force_ontap_version": null,
            "hostname": "<fqdn>",
            "http_port": null,
            "https": true,
            "install_baseline_image": false,
            "key_filepath": null,
            "node": null,
            "ontapi": null,
            "package": null,
            "package_url": "<download url of all_shelf_fw.zip>",
            "password": "VALUE_SPECIFIED_IN_NO_LOG_PARAMETER",
            "reboot_sp": null,
            "reboot_sp_after_download": null,
            "rename_package": null,
            "replace_package": null,
            "server_password": null,
            "server_username": null,
            "shelf_module_fw": null,
            "state": "present",
            "update_type": null,
            "use_rest": "auto",
            "username": "<user_name>",
            "validate_certs": false
        }
    },
    "msg": {
        "job": {
            "_links": {
                "self": {
                    "href": "/api/cluster/jobs/d2840aa2-bb53-11ee-b4bf-d039ea9feb2a"
                }
            },
            "uuid": "d2840aa2-bb53-11ee-b4bf-d039ea9feb2a"
        },
        "job_response": "Firmware update for \"all_shelf_fw.zip\" completed."
    }
}```

#

Audit.log entries:

00000004.00869d9f 0d125ec3 Thu Jan 25 2024 08:23:06 +01:00 [kern_audit:info:2275] 8503ea00001408c5 :: <cluster_name>:http :: DST_IP:33066 :: <cluster_name>:<user_name> :: POST /api/cluster/software/download : {"url":***} :: Success:```




Cluster cli:

``` storage shelf show -fields module-fw-revision,module-latest-fw-revision,module-type
shelf module-type module-fw-revision module-latest-fw-revision
----- ----------- ------------------ -------------------------
3.20  IOM12B      0120,0120          -,-
3.21  IOM12B      0120,0120          -,-
4.10  IOM12B      0120,0120          -,-
4.11  IOM12B      0120,0120          -,-
5.30  IOM12B      0120,0120          -,-
5.31  IOM12B      0120,0120          -,-
6 entries were displayed.

If I download the same package on the cluster cli with storage firmware download it is working

glass heart Jan 25, 2024, 11:47 AM

#

Hey Matthias, I'm not sure if this is the official fix, but yeah, I also stumbled upon this issue.... so applying the ansible script several times (wait 30min between each execution because you do not want to run this task when an upgrade is already in progress) helps 🙂

#

or maybe it's a metrocluster issue, I don't remember exactly, but I think, this behaviour is only shown on metrocluster... and you're probably also running MCC?

eternal field Jan 25, 2024, 11:53 AM

#

I run it yesterday and today. And no change on this.
MCCs and non MCCs behave the same. The update wasn't installed on any of these systems 😦

glass heart Jan 25, 2024, 1:04 PM

#

We will roll out the new shelf firmware as soon as 9.14 is released, so maybe I can use this module within the next 4 weeks, so I can provide an update on this

eternal field Jan 25, 2024, 1:07 PM

#

We are currently on 9.12.Px and plan to update to 9.13.Px in Feburary / March timeframe. 9.14.1 needs at least 5 patch cycles before start to use it :-D.
Updating ~100 HW Clusters and ~100 Select instances takes some time

#

I hope some of teh NetApp guys have an idea, but they are all (most) located on the west coast

lone hedge Jan 25, 2024, 6:42 PM

#

Let me have the team look in to this and get back ro you guys

eternal field Jan 29, 2024, 2:26 PM

#

@lone hedge Could you already have a look at it?

acoustic harbor Jan 29, 2024, 3:00 PM

#

If it works when you download it via CLI, I would use SSH as a workaround and issue the download command that way 🤷‍♂️

lone hedge Jan 29, 2024, 3:40 PM

#

@eternal field it possible the team member who wrote and worked on this in the past is no longer with us so I’ll need to double check.

#Shelf FW download issues