Multiple Canary builds make the Sentry 'New Issues" report inaccurate
Now that we create multiple auto-deploy branches we often have more canary deploys that production deploys.
For example on Sept 7th:
- Staging started (Build A)
- Staging finished (Build A)
- Canary started (Build A)
- Staging started (Build B)
- Canary finished (Build A)
- Canary started (Build B)
- Canary finished (Build B)
- Prod started (Build B)
If we see new Sentry errors from Build A they will not be reported as new issues in Build B's Sentry report. Therefore we either need to check the sentry reports for all builds that have been on canary since the last Prod deployment or find a way to wrap the issues from all un-promoted builds into a single report.