If the Slave Instance is root instance, we should allocated exclusivelly on Subscription defined computers/instances

  Adopt price based on Variations rather them on the lines.

   The goal is define pricing based on the Service Individual Variation in order to support multiple reservation fee prices.

   Introduce a CDN case with different reservation fee and monthly fee pricing.

    extend test to diff movement

This reverts commit 05a8e283

See merge request nexedi/slapos.core!286

lsblk OS inspection is the way to have realiable results for querying the
devices on the system, so fully switch to it, while still supporting symlinks.

Disks and partitions are supported.

Context manager is not needed here and even makes more mess, just simply
patch all what is needed in setUp, then unpatch in tearDown.

   Else simulation behaves oddly.

    This forces to wait for builders before stop

   Payzen has a delay between the Charge and the transference to the bank. This requires the accountant to generate yet a new accounting document to represent it.

     This might replace "Bank" for automation, since "Bank" requires that the money been accurately in relation to Bank Sheets.

    If set, causality is a Subscription Request, and it should be always splited.

   Use shadow user to access Account document, this prevent us to setup complex security.

   The asserts and verification should be done as manager to present set unecessary security roles.

    Those documents are expected to be published when they are accessible

   This isn't compatible with the actuall usage, if we want to keep this, we should install networkcache_erp5 business template (and use groups/security from there).

When slapproxy forwards a request to an external master it is seen as
coming from the slapproxy and the original requester's id is lost.

This means forwarded requests from different requesters can collide.

e.g. two partitions request a frontend and when the second request is
handled the external master updates the first frontend to redirect to
the second partition instead of creating a second frontend.

To avoid collisions, the id of the requester as seen in the slapproxy
is prefixed to the instance name (partition reference) in the request:

- case 1: requester is embedded partition => prefix with partition id
    e.g. 'myinstance' -> 'slappart0_myinstance'

- case 2: requester is user => prefix with 'user'
    e.g. 'myinstance' -> 'user_myinstance'

See merge request nexedi/slapos.core!282