Expected behaviour:

When default_values.puppetclass_ids is defined and the user selects a different puppet environment, the defined Puppet classes keeps automatically assigned to the new host.

Actual behaviour:

The classes are removed and the user needs to select them manually.

After building the package with npm run build and running it with serve -s build, the application returns a white page and the Browser console shows the following error:

index.js:5 Uncaught ReferenceError: process is not defined
    at Module.6 (index.js:5)
    at f ((index):1)
    at Module.1099 (ActionsDropdown.js:9)
    at f ((index):1)
    at Object.425 (lisa-notification-gfx.svg:1)
    at f ((index):1)
    at a ((index):1)
    at Array.e [as push] ((index):1)
    at main.63117b3d.chunk.js:1

We currently use Patternfly v2 and should update it to Patternfly v4.

It should be possible to set a priority for the extension to manage the order in which they are displayed

Foreman and hammer supports oauth/oidc authentication using Keycloak (or any other Idp for that matter). See https://www.theforeman.org/manuals/nightly/index.html#5.7.7SingleSign-onforForemanusingOpenIDConnectprotocol

We should support oauth/oidc auth here as well.

computeResourceId gets initialised to the compute resource for the default location in settings/index.js, however it never gets changed when the location is changed, even though the computeResource itself is.
This results in foreman trying to create a host on a computeResource that might not be available in the alternate location.

My current solution is to simply add

  set(data, 'computeResourceId', computeResource.id)

to hostCreatParams.js, but as I'm not that great with JavaScript this looks like a big hack.

The VM gets created without appTier and domainName; e.g. tst-vm-01 instead of tst-vm-01.dev.location.org.domain

https://vitejs.dev/

The creation of a new VM fails with the following error in the Foreman log:

2020-01-20T16:37:50 [I|app|99294351] Started POST "/api/graphql" for 192.168.182.10 at 2020-01-20 16:37:50 +0100
2020-01-20T16:37:50 [I|app|99294351] Processing by Api::GraphqlController#execute as JSON
2020-01-20T16:37:50 [I|app|99294351]   Parameters: {"_json"=>"[FILTERED]", "graphql"=>{"_json"=>"[FILTERED]"}}
2020-01-20T16:37:50 [I|app|99294351] Authorized user <User>(Manuel Laug)
2020-01-20T16:37:50 [I|app|99294351] Completed 200 OK in 108ms (Views: 3.3ms | ActiveRecord: 19.5ms)
2020-01-20T16:37:50 [I|app|5beede11] Started POST "/api/graphql" for 192.168.182.10 at 2020-01-20 16:37:50 +0100
2020-01-20T16:37:50 [I|app|5beede11] Processing by Api::GraphqlController#execute as JSON
2020-01-20T16:37:50 [I|app|5beede11]   Parameters: {"_json"=>"[FILTERED]", "graphql"=>{"_json"=>"[FILTERED]"}}
2020-01-20T16:37:50 [I|app|5beede11] Authorized user <User>(Manuel Laug)
2020-01-20T16:37:53 [I|app|5beede11] Adding Compute instance for tst-manuel-06
2020-01-20T16:37:53 [W|app|5beede11] Failed to create a compute ESX-LX-LAN (VMware) instance tst-manuel-06: undefined method `resourcePool' for nil:NilClass

EFI should be the default value for RHEL9

It should be possible to make LISA capable to ask for specific/custom fields during the VM creation.

This feature could be used to make sure a specific field in Foreman gets defined for every VM (e.g. host group)

Expected behaviour:
When default_values.puppetclass_ids is defined and the user selects a Puppetserver and environment, the defined Puppet classes get automatically assigned to the new host.

Actual behaviour:
The classes are not selected and the user needs to select them manually.

it seems that pagination does not work and only the first page is visible

If the UI is complete, we should add some screenshots to the README

After successfully creating a VM using LISA, the system does not want to boot and install.
Foreman reports an error when viewing the "NICs" tab on the host:

Failure: undefined method `mtu' for #<Nic::Base:0x00007f6cb1768958>

Comparing the network_interfaces using the Rails console, there seems to be a difference in the interface "type" - see managed vs. base

# created using Foreman GUI (and working)
[
  {
    "managed": {
      "id": 48903,
      "mac": "<MAC>",
      "ip": "<IP>",
      "name": "tst-host-01.apptier.domain.local",
      [...]
      "attrs": {},
      "provider": null,
      "username": null,
      "password": null,
      "virtual": false,
      "link": true,
      [...]
      "managed": true,
      "mode": "balance-rr",
      "attached_devices": "",
      "bond_options": "",
      "primary": true,
      "provision": true,
      "compute_attributes": {
        "type": "VirtualVmxnet3",
        "network": "dvportgroup-1288716"
      },
      "execution": true,
     [...]
    }
  }
]

# created using LISA (not working)
[
  {
    "base": {
      "id": 48871,
      "mac": "<MAC>",
      "ip": "<IP>",
      "name": "tst-host-02.apptier.domain.local",
      [...]
      "attrs": {},
      "provider": null,
      "username": null,
      "password": null,
      "virtual": false,
      "link": true,
      [...]
      "managed": true,
      "mode": "balance-rr",
      "attached_devices": "",
      "bond_options": "",
      "primary": true,
      "provision": true,
      "compute_attributes": {
        "network": "dvportgroup-1288716",
        "type": "VirtualVmxnet3"
      },
      "execution": true,
      [...]
    }
  }
]

As suggested in schema.json there should be some required options, like:

• app_tiers
• form_settings
• default_configs
  (and maybe more)

The build of the application fails because it looks for the options in a different naming scheme (like appTiers).

�[32;1m$ npm run build�[0;m

> [email protected] build /builds/aws/lisa-deployment/lisa
> NODE_PATH=src/ react-scripts build

Browserslist: caniuse-lite is outdated. Please run next command `npm update caniuse-lite browserslist`
Creating an optimized production build...
Failed to compile.

./src/helpers/hostsCreateParams.js
Attempted import error: 'appTiers' is not exported from '../settings'.


npm ERR! code ELIFECYCLE
npm ERR! errno 1
npm ERR! [email protected] build: `NODE_PATH=src/ react-scripts build`
npm ERR! Exit status 1
npm ERR! 
npm ERR! Failed at the [email protected] build script.
npm ERR! This is probably not a problem with npm. There is likely additional logging output above.

The validation of the used settings.json against the schema is successful.

I just noticed that the build of LISA with a plugins fails due to:

$ npm run build
> [email protected] build /builds/lisa
> react-scripts build
Creating an optimized production build...
Failed to compile.
/builds/lisa-plugin/dist/index.js
Cannot find module: '@apollo/react-hooks'. Make sure this package is installed.
You can install this package by running: npm install @apollo/react-hooks.
npm ERR! code ELIFECYCLE
npm ERR! errno 1
npm ERR! [email protected] build: `react-scripts build`
npm ERR! Exit status 1
npm ERR! 
npm ERR! Failed at the [email protected] build script.
npm ERR! This is probably not a problem with npm. There is likely additional logging output above.
npm ERR! A complete log of this run can be found in:
npm ERR!     /home/node/.npm/_logs/2022-04-20T14_32_43_821Z-debug.log

When using the search, the value changes to "[object Object]" and breaks the host view:

For the non-admin user the hosts query only returns records that the user is authorised, so it happens that a name that is already taken is displayed as available

Default value should be "1". This value can be any positive number.

This also requires changes in other parts of LISA.

It seems that deploying two systems simultaneously (using count = 2) results in the first one being successful and the second one failing. Further logs will be collected...

To be able to use Node 12, we need to update node-sass to >= 4.12.0 and maybe more packages.

For reference: sass/node-sass#2662

As puppet is no longer part of Foreman core, we should move puppet stuff to the plugin