Revert "Fixed delays for HTTP retries rather than exponential delays (#1967)"

[pgombar]

Description

Revert #1967
This reverts commit 0991c3a.

---

PR information

• The title of the PR is clear and informative.
• There are a small number of commits, each of which has an informative message. This means that previously merged commits do not appear in the history of the PR. For information on cleaning up the commits in your pull request, see this page.
• Except for special cases involving multiple contributors, the PR is started from a fork of the main repository, not a branch.
• If applicable, the PR references the bug/issue that it fixes in the description.
• New Unit tests were added for the changes made and Travis.CI is passing.

Quality of Code and Contribution Guidelines

• I have read the contribution guidelines.

[cjp256]

This was many moons ago, but I wonder if anyone remembers why this was reverted? :-) I have a case where it seems that provisioning timed out in this retry loop. We can adjust the provisioning logic to not use retries, but I am curious why such a significant backoff. Provisioning has 30 retries with 15s retry delay which IIUC leads to:

delay for attempt 1: 30
delay for attempt 2: 45
delay for attempt 3: 75
delay for attempt 4: 120
delay for attempt 5: 195
delay for attempt 6: 315
delay for attempt 7: 510
delay for attempt 8: 825
delay for attempt 9: 1335
delay for attempt 10: 2160
delay for attempt 11: 3495
delay for attempt 12: 5655
delay for attempt 13: 9150
delay for attempt 14: 14805
delay for attempt 15: 23955
delay for attempt 16: 38760
delay for attempt 17: 62715
delay for attempt 18: 101475
delay for attempt 19: 164190
delay for attempt 20: 265665
delay for attempt 21: 429855
delay for attempt 22: 695520
delay for attempt 23: 1125375
delay for attempt 24: 1820895
delay for attempt 25: 2946270
delay for attempt 26: 4767165
delay for attempt 27: 7713435
delay for attempt 28: 12480600
delay for attempt 29: 20194035

FWIW it is perhaps worth reconsidering the name "retry_delay" as it's more an initial delay multiplied by 2.