Presented here is a subset of metrics, notifications and alerts that the management infrastructure of Sonic ESB exposes.
This subset is a basic and critical list to be included in your watchlist so as to not miss any event that potentially indicates a failure in the system and therefore requires prompt action.
The ones included here indicate fault and offline conditions.
Fault conditions, when notified, should always be looked into. Offline conditions should be investigated when not expected (ie when the component is not deliberately shutdown).
Gathering of metrics, notifications and alerts places additional burden on the management infrastructure and the component being watched.
Therefore it is important to restrict the amount of monitoring. If the performance impact is negligible, or if the requirement is for restricted periods of time, additional metrics, notifications and alerts may be included. Message count on queues (that feed into key services, dead message queues and routing queues, when the Dynamic Routing Architecture is used) is one metric of common interest.
Which version of Sonic does this document cover ?Thanks.