Honor collector_database_exclusions in default trace and trace analysis collectors (#887 follow-up)

[erikdarlingdata]

Summary

• Follow-up to PR Per-database exclusions for collectors — Dashboard side (#887) #905. default_trace_collector and trace_analysis_collector were skipped in the original per-database exclusion work because they filter trace events by DatabaseID / DatabaseName rather than iterating sys.databases, so the NOT EXISTS (... d.name) pattern from Per-database exclusions for collectors — Dashboard side (#887) #905 didn't drop in directly.
• Both collectors now honor config.collector_database_exclusions, matching the behavior of the other 8 per-database collectors.

Changes

• install/29_collect_default_trace.sql — NOT EXISTS against config.collector_database_exclusions joined to sys.databases to translate the exclusion list (by name) into ids matching ft.DatabaseID from sys.fn_trace_gettable.
• install/31_collect_trace_analysis.sql — NOT EXISTS directly on trc.DatabaseName (the .trc file exposes the name, no translation needed).

Scope check

Re-audited every Dashboard install/*collect*.sql and every Lite collector class. Confirmed:

• These two were the only Dashboard collectors missing the filter.
• Lite is fully covered — all 9 per-database collectors apply BuildDatabaseExclusionFilter() from RemoteCollectorService.cs. No Lite changes needed.

Test plan

• Run the CLI installer end-to-end against a test instance and confirm 54/54 install scripts succeed.
• Verify both procs reference config.collector_database_exclusions after install.
• Smoke test: exclude a database that's currently producing trace/audit events (e.g. distribution), confirm default_trace_collector and trace_analysis_collector skip its events while a non-excluded DB still produces rows. Wrap in a transaction and roll back.

Addresses the follow-up question on #887 (comment).