Troubleshooting, Governance, and Best Practices
Troubleshooting questions reward a systematic approach. A candidate must know where to look: logs (error.log, request.log, access.log), the OSGi console and bundle states, the Sling servlet resolution, and the dispatcher logs. Understand common failure patterns: a component not rendering (resource type mismatch, missing model, HTL error), a page not caching (headers, rules), a 403 (permissions or filter), and a 404 (mapping or replication). Know how to use the query performance tool and the JCR observation and indexing diagnostics.
Governance concepts include coding standards, the use of core components and proxy patterns, package structure, and the review process. Understand the role of Cloud Manager quality gates, code coverage thresholds, and the importance of automated tests. Know the content governance model: workflows for approval, versioning, and the separation of duties between authors and approvers. Also understand the value of naming conventions for templates, components, and policies.
Traps include jumping to a fix without isolating the layer (author, publish, dispatcher, CDN), or assuming a local success implies production success. Another trap is ignoring the difference between a content issue and a code issue. Expect questions on why a change did not appear on publish (replication queue, flush agent, cache) and on how to diagnose slow pages (requests, queries, client libraries).
Rule of thumb: isolate the layer first, then the artifact. Worked example: a page renders on author but 404s on publish. Check replication status, then the dispatcher filter, then the resource resolver mapping, in that order.