IDD 5.0 review discussion - Orchestration-pull

[AlexChiquito]

n this Issue we will collect the comments about the orchestration-pull-http-json interface definition.

[emanuelpalm]

Link to reviewed document: eu.arrowhead.orchestration-pull-http-json.yml.

[borditamas]

AITIA review comments

• File name should be 'service-orchestration'
• In case of orchestration there is an inconsistency between the IDDs and the SysD. In the SysD there is one orchestration service with three operations. One for pull and two (subscribe, unsubscribe) for the push orchestration. In the the IDDs, the pull and push are different services. During the previous roadmap discussions it was decided to have only one orchestration service.
• This IDD should state that only describes the store orchestration strategy.

---

https://github.com/eclipse-arrowhead/roadmap/blob/5a4f4b2d3149288a48bda69337f7e041a05d0161/5.0%20Draft/IDD/IDDs%20Orchestration%20System/eu.arrowhead.orchestration-pull-http-json.yml#L25

"echo" operation should not be part of this service, but a different "monitor" service.

---

https://github.com/eclipse-arrowhead/roadmap/blob/5a4f4b2d3149288a48bda69337f7e041a05d0161/5.0%20Draft/IDD/IDDs%20Orchestration%20System/eu.arrowhead.orchestration-pull-http-json.yml#L34

The phrase "recommendation" shouldn't be used in the operation. It is up to the developer of the consumer system that it is considering the orchestration results as a recomenndation or not.

---

https://github.com/eclipse-arrowhead/roadmap/blob/5a4f4b2d3149288a48bda69337f7e041a05d0161/5.0%20Draft/IDD/IDDs%20Orchestration%20System/eu.arrowhead.orchestration-pull-http-json.yml#L34

• We don't understand the persence of the auth-tokens in the orchestration request. This orchestration service should never use tokens (and if uses, it should be interface independent and should be placed in the header in case of HTTP).

• The input payload is way too simple-store strategy specific. It should cover all the three strategies we have discussed about: simple-store, flexible-store, dynamic. Also, the output payload should cover the case when the consumer gets all the access info as well (not just service instance ids).

[borditamas]

@AlexChiquito @emanuelpalm @PerOlofsson-Sinetiq Could you please provide Sinetiq's feedback before the next RoadMap (05.02) in order to being able to discuss it there? As you know, last time the 14th of May (before AIMS 5.0 GA) was agreed to target the specification being finalized, so we don't have so much time.