Know what is wrong with your database in five working days.
A senior DBA reviews your SQL Server or PostgreSQL instance from end to end. You get a written report with every finding ranked by risk and effort, scripts for the quick fixes, and a call to walk you through it.
- Report in 5 working days
- Remote, read-only access
- No changes to your systems
When it makes sense
- You still run SQL Server 2016 or 2017. 2016 left extended support in July 2026, 2017 follows in October 2027.
- Your PostgreSQL is version 13 or 14. Both are at or near end of life.
- Users complain about slow screens, timeouts or night jobs that run into the morning.
- Nobody has tested a full restore in the last year.
- You face a NIS2, DORA or internal audit and need evidence that backups and access are under control.
- Licence costs keep rising and you want to know whether PostgreSQL is a realistic option.
What we check
More than 80 checks, grouped by area. The same method for both engines, adapted to each one.
SQL Server
- Version, patch level and support dates
- Instance and database configuration (MAXDOP, memory, tempdb, files)
- Wait statistics and the heaviest queries from Query Store
- Missing, unused and duplicate indexes
- Always On AG or failover cluster health and failover readiness
- Backup chain, RPO/RTO and a real restore test
- Logins, sysadmin rights, encryption (TDE, TLS) and auditing
- SQL Agent jobs, maintenance plans and error logs
PostgreSQL
- Version, minor releases and extensions
- postgresql.conf: memory, WAL, checkpoints, connections
- pg_stat_statements: the heaviest and most frequent queries
- Indexes, table bloat and autovacuum behaviour
- Replication, Patroni and failover readiness
- Backups with pgBackRest or Barman, PITR and a restore test
- Roles, pg_hba.conf, SSL and row-level security
- Connection pooling (PgBouncer) and lock contention
MySQL, MariaDB and MongoDB on request.
What you receive
- Executive summary
- One page for management: overall status, the three biggest risks and what they cost if nothing changes.
- Findings, ranked
- Every finding with severity, evidence from your system, the fix and the effort it takes.
- Fix scripts
- T-SQL or SQL scripts for the quick wins, commented and ready for your team to review and run.
- Roadmap
- What to do this week, this quarter and this year, in that order.
- Walkthrough call
- 60 minutes with the DBA who wrote the report, to answer your team's questions.
| High | Backup | Log backups run every 4 hours: up to 4 hours of data loss on a system that needs 15 minutes. |
|---|---|---|
| High | Performance | 38% of wait time is PAGEIOLATCH_SH. Three missing indexes on dbo.Orders cause full scans. |
| Medium | High availability | The synchronous replica in the second site adds 12 ms to every commit. |
| Medium | Security | Four logins with sysadmin rights are not used by anyone. |
| Low | Configuration | MAXDOP and cost threshold for parallelism are still on default values. |
Vincenzo Petrucci, founder of Pacific IT Solutions. Senior DBA and T-SQL developer with more than twenty years on critical SQL Server and PostgreSQL systems for banks, insurers and telecom operators. He writes every report himself.
About →How it works
Kick-off
A 30-minute call. We agree the scope and you create a read-only login with a script we provide.
Data collection
Our scripts read configuration, statistics and query history. Nothing is changed or installed.
Analysis
A senior DBA reviews the data and confirms each finding before it goes into the report.
Report and call
You receive the report on day 5 and we walk through it together.
Safe for production
Read-only
We use the minimum monitoring permissions: VIEW SERVER STATE on SQL Server, pg_monitor on PostgreSQL.
Your data stays with you
We read metadata and statistics, never the content of your tables.
NDA on request
We sign your non-disclosure agreement before we start.
Questions
Do you need admin rights?
No. A read-only monitoring login is enough. We send you the script to create it.
Will it slow down production?
No. Data collection uses lightweight system views and runs in a few minutes.
What counts as one instance?
One SQL Server instance or one PostgreSQL cluster. Replicas in the same Availability Group or Patroni cluster are included.
Can you fix the problems too?
Yes, as a follow-up project or monthly retainer. If you order follow-up work within 60 days, the price of the health check is deducted.
Which versions do you cover?
SQL Server 2012 to 2025 and PostgreSQL 12 to 18, on-premise, Azure, AWS or containers.
Book your health check
Write to us with the engine, the version and where it runs. We reply within one working day with a start date.