NAV 2009 DB Maintenance

Targets the SQL Server database behind a Dynamics NAV 2009 install. The bundled script scripts/Invoke-NavDbMaintenance.ps1 plans maintenance and (only with -Execute) runs it, emitting JSON throughout. Default mode is a DRY RUN — the script prints the exact T-SQL it would execute and changes nothing without -Execute. It only rebuilds/reorganizes and updates statistics on existing indexes; it NEVER creates or drops indexes (NAV owns them). It never calls an LLM.

Before running maintenance, diagnose the state first with nav2009-sql-performance (fragmentation, stale stats, SIFT sizes). Then plan an action here and confirm the T-SQL before executing.

SCRIPT = this skill's scripts/Invoke-NavDbMaintenance.ps1. Requires PowerShell 7+ (pwsh) and network access to the SQL Server.

Permissions & auth

• Default is Windows integrated auth (the user running pwsh). Pass -SqlCredential for SQL auth — it is a PSCredential; let PowerShell prompt (-SqlCredential (Get-Credential)). Never put a password on the command line.
• Required permissions by action:
  • backup — db_backupoperator (or db_owner)
  • index_maintenance — db_owner or ALTER permission on the indexes
  • checkdb — db_owner or sysadmin
  • statistics — db_owner or UPDATE STATISTICS permission
• Connections are unencrypted by default (NAV 2009-era instances rarely have TLS certs); add -Encrypt if the instance supports it.

How to run

Always run with pwsh. Without -Execute the script is a dry run — it returns planned T-SQL in each section's plan array and sets executed: false.

Want	Pass
Full maintenance plan (dry run)	-ServerInstance SQLSRV01 -Database NAV_PROD
Just back up	-Actions backup
Just index maintenance	-Actions index_maintenance
Just integrity check	-Actions checkdb
Just update statistics	-Actions statistics
Change fragmentation thresholds	-ReorganizeThreshold 15 -RebuildThreshold 25
ONLINE rebuild (Enterprise only)	-Online
Actually execute	add -Execute
SQL auth	-SqlCredential (Get-Credential)
Save report	-OutFile C:\ops\maint-report.json

# Dry run — see the full maintenance plan before committing
pwsh -File SCRIPT -ServerInstance SQLSRV01 -Database 'NAV_PROD'

# Execute a backup only (use a maintenance window for index rebuild / CHECKDB)
pwsh -File SCRIPT -ServerInstance SQLSRV01 -Database 'NAV_PROD' `
    -Actions backup -BackupPath 'D:\Backups\NAV_PROD.bak' -Execute

Output contract

• Without -OutFile → full JSON on stdout.
• With -OutFile → full JSON to the file; a compact summary (per-section status) on stdout. Prefer -OutFile for large runs so your context stays small; then read only the sections you need.

Top level: status (ok / partial / error), generated_at (UTC ISO-8601), executed (bool), and sections keyed by action name. Each section:

{
  "backup":            { "status": "planned|ok|error|skipped", "plan": ["T-SQL..."], "data": {} },
  "checkdb":           { "status": "planned|ok|error|skipped", "plan": ["T-SQL..."], "data": {} },
  "index_maintenance": { "status": "planned|ok|error|skipped", "plan": ["ALTER INDEX..."], "data": {} },
  "statistics":        { "status": "planned|ok|error|skipped", "plan": ["EXEC sp_updatestats"], "data": {} }
}

In dry run (-Execute absent): each runnable section has status: planned and a plan array of the exact T-SQL statements — no database state is changed. With -Execute: statements run, status becomes ok or error, and data reports the outcome (backup file size, CHECKDB result, indexes rebuilt/reorganized, stats updated).

What you (the agent) do with the result

1. Dry run first, always. Run the script without -Execute, show the user the planned T-SQL from each section, and get explicit confirmation before re-running with -Execute — especially for CHECKDB and index rebuild on a production DB during business hours.
2. Recommend a maintenance window. CHECKDB and index rebuild hold locks and can be I/O-heavy. Suggest running outside peak NAV posting hours. CHECKDB on a large DB can run for hours.
3. After execution, report what changed: which indexes were rebuilt vs reorganized (table name, index name, fragmentation before), backup file path and size, CHECKDB pass/fail with any messages.
4. NEVER suggest creating or dropping indexes. If fragmentation persists after maintenance it is a NAV key or SIFT design issue — point the user to nav2009-development (to adjust key design) and nav2009-sql-performance (to see which SIFT views or indexes are candidates for MaintainSIFTIndex/MaintainSQLIndex off).
5. After execution, remind the user that wait stats and index-usage stats reset on SQL Server restart — re-triage with nav2009-sql-performance afterward to see the new baseline.

Errors

• Cannot connect — wrong instance name, SQL Browser off for named instances, firewall, or auth mismatch. Check with Test-NetConnection <server> -Port 1433.
• Permission denied on backup — the SQL service account must have write access to the backup path; the login needs db_backupoperator or db_owner.
• ONLINE rebuild unsupported — ALTER INDEX ... REBUILD WITH (ONLINE=ON) needs Enterprise edition; on Standard/Web/Express the section will error with "ONLINE is not supported". Remove -Online and schedule an offline rebuild in a maintenance window.
• CHECKDB permission denied — needs db_owner or sysadmin. The other actions still run (each section is independent).
• Backup path not writable by SQL service account — the path is resolved on the SQL Server, not the client machine. Ensure the SQL Server service account has write access to the UNC/local path specified.