Breakerbox 0.3.3

• Dropwizard 0.9.2
• Tenacity 0.7.3
• Postgresql 9.4.1207.jre7

Breakerbox 0.3.2

• Tenacity 0.7.2
• Merge dashboard fixes Netflix/Hystrix#959

Breakerbox 0.3.1

• Dropwizard 0.9.1
• Tenacity 0.7.0
• dropwizard-auth-ldap 0.3.0
• Upgraded maven-plugins
• Postgresql 9.4-1205-jdbc42
• H2 1.4.190

Breakerbox 0.2.12

• Tenacity 0.6.16
• Fixing dashboard issue where when not all circuits were open it could cause the circuit to not render properly.

Breakerbox 0.2.11

• Tenacity 0.6.15
• Merge dashboard fixes Netflix/Hystrix#919
• Merge dashboard fixes Netflix/Hystrix#921

Breakerbox 0.2.10

• Tenacity 0.6.14
• H2 1.4.189
• Postgresql 9.4-1203-jdbc42
• Fixes a bug in config.properties it would not respect an instance:port:path when referencing an instance. It would only honor the instance:port and not the path.
  Now if you specify 10.0.0.1:8080/some/root, it will look for tenacity resources at 10.0.0.1:8080/some/root/tenacity/...

Breakerbox 0.2.9

• Dropwizard 0.8.4
• Postgresql 9.4-1202-jdbc42
• Tenacity 0.6.13
• Dropwizard-auth-ldap 0.2.1

Breakerbox 0.2.8

• Tenacity 0.6.12
• Only advertise Archaius values for executionIsolationStrategy if it was explicitly set by a user. This allows for the specific type of command TenacityObservableCommand or the original TenacityCommand to determine
  what kind of isolation strategy to use. Only override this behavior if the user sets it on their own.

Breakerbox 0.2.7

• Tenacity 0.6.10
• Adding Bad Request Counts to the dashboard

Breakerbox 0.2.6

• Tenacity 0.6.9
• Switch from using latency_execute to latency_total for percentile metrics. The latency_execute is a closer measurement to successful execution times. Where latency_total is more accurate at the actual time the application is experiencing for successful and timed out commands. Since we assume breakerbox is used more in a debugging response and investigations rather than profiling/benchmarking, latency_total is more helpful in those quick response situations. In comparison when evaluating if your threadpool sizes and timeouts are set appropriately it is better to look at a historical trend of the latency_execute metric.