A job running in AWX prevents an inventory sync (of the same
inventory). This restricts AWX to run only one job at a time as
the inventory is always updated on launch.
To prevent this issue, the inventory is updated from AWX when needed.

- The update is triggered on any model modification that is used in the
inventory (AnsibleGroup, Cname, Domain, Host, Network, NetworkScope)
- For Interface, the inventory sync is triggered by the
trigger_core_services function as it depends on it
- The inventory update job is put on the "high" queue so that it takes
precedence over all other jobs.

JIRA INFRA-887 #action In Progress

Ensure that the trigger_core_services_update job is always processed
before the other jobs (low queue).

JIRA INFRA-887

- users with access to the network can create VIOC
- VM can be created only in the ALLOWED_VM_CREATION_DOMAINS

JIRA INFRA-775

- vm_osversion is passed as extra variables to the create_vm job
template
- vm_cores/vm_memory/vm_disk_size are also passed as extra variables
instead of saving them as host variables. Saving the variables in the
host when submitting the form forces to update the inventory before to
trigger the job to create the VM. This can be avoided using extra
variables. Furthermore those variables don't need to be in the
inventory, they can be retrieved as facts.

JIRA INFRA-877

Only vmname and domain are required for the ics-ans-deploy-on-proxmox
playbook. All other variables are taken from the inventory.

JIRA INFRA-870 #action In Progress

New resource argument can be passed to launch_job_template.
It is optional and is set to "job" by default.
Shall be set to "workflow_job" to run a workflow template.

Note that passing the workflow name didn't work during my test.
The id had to be used instead.

JIRA INFRA-867 #action In Progress

After a VM or VIOC creation, a post install job can be triggered.
The job to trigger is based on the domain and defined in the
AWX_POST_INSTALL dictionary.

Admin users have the option to skip this post job if required.

The job is enqueued in RQ with a dependency on the creation job.
So it's only triggered after successful completion of the first one.

JIRA INFRA-769

- Disk size added to the create VM form
- Disk size saved to proxmox_disk_space variable in host inventory

The proxmox_disk_space variable is used by the Ansible job when creating
the machine.

JIRA INFRA-759 #action In Progress

- flask-bootstrap dependency removed a while ago
- the full static directory shall be passed

JIRA INFRA-683

JIRA INFRA-578 #action In Progress

JIRA INFRA-578

dns parameter removed (retrieved dynamically by the playbook)

JIRA INFRA-651 #action In Progress

JIRA INFRA-575 #action In Progress

JIRA INFRA-575

Retrieving all items from the database takes some time.
The task is thus handled by the RQ worker.
We use ajax to periodically monitor the status of the job
and redirect to the file when ready.
A progress bar is displayed to let the user see the progress.

The mechanism put in place could be re-used for other use cases.

JIRA INFRA-545 #action In Progress

WARNING! To pass the SECRET_KEY as en env variable, you should use
base64 encoding as the key is supposed to be a random string of bytes.

Errors are now sent to sentry.
No need to send them directly via e-mail.

JIRA INFRA-525

JIRA INFRA-525 #action In Progress

CSENTRY_ENVIRONMENT shall be set to production|staging|development
It can be used when sending errors to Sentry

JIRA INFRA-525

JIRA INFRA-414

https://black.readthedocs.io/en/stable/

Change "CSentry" to "CSEntry-test" in the navbar (upper left corner)
and set it to red on the staging server.

This is done by setting the variable CSENTRY_STAGING to True in the
settings.cfg file for the staging server (variable set to False by
default).

JIRA INFRA-347

Both AWX_VM_CREATION_ENABLED and AWX_JOB_ENABLED shall be set to True
to enable VM creation.
An extra variable is used because the core services can be updated from
csentry-test but VM creation shouldn't (for now)

Allow to create a VM or Virtual IOC

A new template 'ics-ans-core @ DHCP test' was created to test
the DNS and DHCP deployment. This is the one o use by default.

The "real" template should be used in the production settings.cfg

AWX jobs should only be triggered on the production server (not from
staging). At least until we get a full staging environment.

- RQ Dashboard shall only be accessible to admin users
- "REDIS_URL" is hard-coded in rq-dashboard (had to rename RQ_REDIS_URL)

The "REDIS_URL" variable is hard-coded in RQ Dashboard.

REDIS_URL was used by redis_store that is passed to Flask Session.
The url variable used by Flask-Redis can be changed by setting the
config_prefix.

1. Rename REDIS_URL to SESSION_REDIS_URL
2. Rename redis_store to session_redis_store
3. Use "SESSION_REDIS" config_prefix to initialize session_redis_store

This makes it clear that this Redis instance is used by Flask-Session.
REDIS_URL will be used for RQ.

Not everyone at ICS is part of the 'ICS Employees' AD group.
We have to accept 'ICS Consultants' as well.

The CSENTRY_LDAP_GROUPS settings now takes a list of groups instead of
just one.

Remove unused is_member_of_all_groups method.

Fixes INFRA-225

Fixes INFRA-224

Fixes INFRA-197

It shouldn't be possible to change the ICS id of an item.
But when registering a bunch of existing items via the API, we need to
allocate an ics_id otherwise not much can be done in the interface.
The id is required for the view and edit page.

The solution is to define some "temporary" ids that don't correspond to
any pre-printed labels.
Items with a temporary id can be assigned a real ics_id when assigning a
label.

Reserved ids start with 'ZZ' (from ZZA000 to ZZZ999)

ICS id can now be non nullable.

Allow to customize the default selected values:
- the default network prefix can be defined in the configuration (default to 24)
- the default first and last IP chosen can probably be improved
  it defaults to hosts[9] and hosts[-5] if the list has more than 15
  elements or the first and last otherwise

Note that when using a small prefix (< 19), the number of possilbe IPs is quite
large and updating the first and last IP select field takes some time...

In Flask, session data are stored in a cookie by default.
Using a server side session, there is no risk of putting too much data
in the session and transmissting a big cookie on every request
(only an id is sent in the cookie).

Redis is used as backend (without persistent storage).

- All created tokens are stored in the database.
  To revoke a token, we just delete it from the database.
  Tokens not found in the database are thus considered unvalid / revoked.

- Add button to copy generated token to the clipboard

- Redesign profile page