Best practice 1 : Retry

[barbara79]

In order to add proper retries to the sdk, we can try to implement retry middlewares with different retry strategies. prefrably make a PR per middleware to keep focus on each case.

create minimum 5 Middleware one for each type of retry:

• Exponential back-off
• Incremental intervals
• Regular intervals
• Immediate retry
• Randomization

see here for technical details https://docs.microsoft.com/en-us/azure/architecture/best-practices/transient-faults

linked to #192

[barbara79]

there is already an Exponential back-off