GitLab

Was Revert by Seb:

Using the public cloudooo for all unit tests makes things unreliable, we have every day many tests failing due to this. It would be required to reapply this change to :

• make ERP5 more resilient to network failures when there is issues connecting to cloudooo
• find some automated ways to have a cloudooo per testnode machine, with idea of having possibility to request "where is the nearest service doing that"
• possibly having a CDN of cloudooo, but this does not prevent having a cloudooo per testnode

---

This MR provides two commits :

• Remove Cloudooo from stack/erp5 ;
• Add parameter to configure Cloudooo URL.

By default, ERP5 will be configured (at instanciation) to use the public cloudooo cluster on https://cloudooo.erp5.net/ (defined on Default System Preference).

But you can choose which URL to use. Here is an example of instanciation parameter : {"cloudooo-url":"https://example.com/"}

caution /!\ as concequences Cloudooo will become orphean if you upgrade your site from a slapos computer or from a webrunner :

• from a slapos computer, the Cloudooo instance won't be requested anymore by the root instance. So here would be the suggested steps to upgrade :
  • run slapos node software to build the new software ;
  • configure Cloudooo URL on your site to point to an external Cloudooo ;
  • stop Cloudooo service ;
  • request Cloudooo stop (or destroy) ;
  • once software built, run slapos node instance.
• from webrunner, the differences is that it's not possible to destroy (or stop?) the Cloudooo instance. Here are the suggested steps to upgrade :
  • configure Cloudooo URL on your site to point to an external Cloudooo ;
  • stop Cloudooo service ;
  • then hack to make slapos proxy to think Cloudooo does not exist anymore (stopped or destroyed) ;
  • click Play.