We use a combination of tools to ensure that Hub-V2:
- Can support at least 150% of the current peak production load
- Guarantee ordering for Sequential Write Use Case
- Ensure that all items are written from Spoke into S3
Hourly Rolling Restart
Every hour a jenkins job performs a rolling restart of the Hub-V2 instances to expose any issues with a deploy or rolling restart.
S3 Write Behind Verification
A jenkins job runs every half hour to make sure that all items in the Hub’s Spoke are also in S3. It executes a jasmine spec