[CURATOR-441] Recipes such as TreeCache are susceptible to herding on reconnections

• Type: Improvement
• Status: Closed
• Priority: Minor
• Resolution: Won't Do
• Affects Version/s: 4.0.0
• Fix Version/s: TBD
• Component/s: Framework, Recipes
• Assignee: Jordan Zimmerman
• Reporter: Jordan Zimmerman
• Created: 08/Nov/17 02:29
• Updated: 22/Apr/20 03:37
• Resolved: 22/Apr/20 03:37

Description

Many of Curator's recipes (TreeCache, etc.) have to refresh their state when a connection problem is repaired. In larger systems, this could cause a thundering herd as all these clients press the ZK ensemble. It would be nice to introduce an optional Jitter mechanism into the Framework connection management so that reconnections aren't immediately reported to the recipes, but are instead delayed slightly with some randomness.

Issue Links

• is superceded by: CURATOR-533 Improve CURATOR-505 by making the CircuitBreaker instance shared