eb4b711679
This CR adds the remaining pieces need to allow worker task processing to add tasks needing retries to the order retry tasks table, and then to have the periodic task service pick these retry tasks up and enqueue them for retries. With this initial CR no attempt is made to protect against concurrency between the existing worker RPC process and the separate periodic retry scheduler process. A subsequent CR will add concurrency protection. This CR does allow the retry behavior to be demonstrated locally as follows: 1) run the 'barbican.sh start', then 2) run the new barbican-worker-retry-scheduler.py script, then 3) POST to the orders resource a 'certificate' type order, and finally poll this new order until it becomes ACTIVE. The simple certificate plugin will call for a retry for the status check method after the initial certificate request is made. The retry processing will eventually call the plugins status check method, which then produces a fake certificate. Change-Id: Idee08abcd3bf2920039f91a7f5b6125eacd1f785 |
||
|---|---|---|
| .. | ||
| barbican | ||
| init | ||
| logrotate.d | ||