feat(sdlc): add indexing errors logs and remove duplicated debug logs

What does this MR do and why?

Improves logging in the SDLC indexer module for better production debugging and monitoring.

• Failure logs (warn -> error): watermark fetch failures, pipeline failures, and lock release failures now log at error level so they show up in alerts.
• Elapsed time tracking: global and namespace handlers now measure and log how long indexing runs take, down to individual entity and edge pipelines. Useful for spotting slow pipelines.
• Less log noise: removed debug logs that just echoed "starting pipeline" / "pipeline completed" without adding anything. Pipeline completions with zero rows are now debug instead of info.
• Log query params instead of query text: the extract queries are static SQL templates, so logging them every run is pointless. The params (watermark ranges) actually change between runs.
• Flattened control flow in namespace handler: continue after errors instead of nesting the success path in else blocks.
• Structured error summaries: when indexing finishes with failures, a dedicated error log reports failed vs. successful pipeline counts with elapsed time.

important note: Metrics and dashboards are coming in in a follow-up merge request.

Performance Analysis

• This merge request does not introduce any performance regression. If a performance regression is expected, explain why.