SQL Server consulting · Performance, Always On, Azure SQL, upgrades, security · 24×7 remote DBA
SQL Server Consulting Measured in Wait Statistics, Plan Regressions and Restore Times, Not Opinions
MinervaDB SQL Server consulting is delivered by senior engineers who have run SQL Server in production since the 2005 era, across on-premises estates, Always On availability groups, Azure SQL Database and Managed Instance, and Amazon RDS. Every recommendation names the DMV, Query Store view or drill result that justifies it; every change ships with a blast radius and a rollback written before execution; and the same team that designs the platform carries the 24×7 watch afterwards. Vendor-neutral by principle: we sell no licences, so when per-core licensing no longer earns its keep, we say so.
01 · Why MinervaDB for SQL Server consulting
SQL Server rewards discipline and punishes defaults
SQL Server ships with a mature optimizer, Query Store, Always On and a rich DMV surface. It also ships with defaults, MAXDOP, cost threshold, tempdb layout, auto-growth, that were never meant for the workload you run, and a licence model that turns every unnecessary core into a recurring cost.
Vendor-neutral
No licences sold, no Azure resale margin, no partner incentives. SQL Server consulting recommendations are based on what your wait statistics and invoice say, including when the right answer is Azure SQL Managed Instance, a smaller edition, or a staged exit to PostgreSQL.
Senior engineers only
Every engagement is led by principal-level SQL Server DBAs who have handled parameter-sniffing regressions at month-end, quorum loss on a Friday night and log chains broken by a well-meaning script. No junior bench learning on your time.
Real 24×7 operations
Follow-the-sun coverage across APAC, EMEA and the Americas with a senior engineer on watch, S1 acknowledged within 15 minutes, every action recorded with timestamps in a shared incident system. SQL Server 24×7 support →
Measurement first
In SQL Server consulting at MinervaDB nothing is claimed as an outcome until the number has moved: wait profile, duration and CPU per query hash, redo queue, restore time, licence cost per core. Estimates are labelled as estimates.
02 · SQL Server consulting services
Six disciplines around the production estate, one accountable team
Most SQL Server consulting engagements start with the performance audit and grow into the discipline the evidence points at. All six are delivered by the same engineers.
Figure 1. The MinervaDB SQL Server consulting service map: performance, HA and DR, migration, upgrades and lifecycle, security and compliance, and licensing and cost, entered through the performance audit and sustained by 24×7 remote DBA operations.
Performance engineering
The core of SQL Server consulting: Query Store-driven tuning, wait-statistics analysis, index and statistics engineering, plan forcing and hints used as scalpels, tempdb and memory-grant contention, parameter-sensitivity settings on 2022 and later, and configuration changes stated as parameter, current versus proposed value and restart requirement.
High availability and disaster recovery
Always On availability groups and distributed AGs, failover cluster instances, log shipping where it still fits, WSFC quorum design with cloud witness, listener and read-only routing, and DR drills that produce measured RPO and RTO rather than design-document targets.
Azure SQL and cloud migration
Assessment with Data Migration Assistant and Azure Migrate, landing-zone selection between Azure SQL Database, Managed Instance, SQL Server on VMs and Amazon RDS, Managed Instance link and distributed AGs for near-zero-downtime cutovers, and Azure Hybrid Benefit and reserved capacity modelled from utilisation. Cloud database FinOps →
Upgrades and version lifecycle
Exits from 2014 and 2016 (both out of support), planned exits from 2017 and 2019, upgrades to 2022 and 2025 by AG rolling upgrade or side-by-side migration, compatibility level raised only after Query Store regression review, and a cumulative-update calendar operated for you.
Security and compliance
TDE with customer-managed keys, Always Encrypted with secure enclaves, TLS enforcement, least-privilege role design, row-level security and dynamic data masking, SQL Server Audit retained for the compliance window, and the evidence pack for GDPR, HIPAA, PCI DSS and SOC 2.
Remote DBA and 24×7 support
Senior SQL Server DBAs operating your estate: monitoring tuned to the workload, incident response with root-cause analysis, patching and CU management, backup verification with timed restore drills, index and statistics maintenance that does not fight your workload, and a monthly SLO report. Remote DBA services →
03 · SQL Server consulting method
Baseline, attribute, change, validate: how the work is measured
SQL Server tells you what it is waiting on. The method is to listen to that before touching anything, change one thing, and re-measure.
Figure 2. The SQL Server performance engineering method and the evidence surface behind it: which DMV, Query Store view or feature proves each class of finding.
A SQL Server consulting baseline is captured over a window that includes month-end and batch peaks: sys.dm_os_wait_stats deltas, Query Store top resource consumers by duration, CPU and logical reads, blocking chains from sys.dm_exec_requests, memory grants pending, I/O stalls per file from sys.dm_io_virtual_file_stats, tempdb allocation-page latch waits, and AG send and redo queues. Attribution names the constraint from the wait class and the plan shape, with the query hash that reproduces it.
Changes are made one at a time and are reversible: an index with a documented drop, a statistics update with persisted sample rate, a forced plan that can be unforced, MAXDOP and cost threshold for parallelism set from the wait profile, additional tempdb files, a compatibility-level change staged behind Query Store. Blast radius and rollback are written before execution.
-- Wait profile since last restart, excluding idle waits
-- (delta this against a snapshot taken at the start of the window)
SELECT TOP (15)
wait_type,
waiting_tasks_count,
wait_time_ms / 1000.0 AS wait_s,
signal_wait_time_ms / 1000.0 AS signal_s,
100.0 * wait_time_ms / SUM(wait_time_ms) OVER () AS pct
FROM sys.dm_os_wait_stats
WHERE wait_type NOT IN (
'SLEEP_TASK', 'BROKER_TO_FLUSH', 'CHECKPOINT_QUEUE',
'LAZYWRITER_SLEEP', 'XE_TIMER_EVENT', 'DIRTY_PAGE_POLL',
'LOGMGR_QUEUE', 'REQUEST_FOR_DEADLOCK_SEARCH', 'WAITFOR',
'SP_SERVER_DIAGNOSTICS_SLEEP',
'HADR_FILESTREAM_IOMGR_IOCOMPLETION')
ORDER BY wait_time_ms DESC;
-- Query Store: slowest plans in the last 24 hours
SELECT q.query_id, p.plan_id,
SUM(rs.count_executions) AS execs,
AVG(rs.avg_duration) / 1000.0 AS avg_ms,
AVG(rs.avg_logical_io_reads) AS avg_reads
FROM sys.query_store_runtime_stats rs
JOIN sys.query_store_plan p ON p.plan_id = rs.plan_id
JOIN sys.query_store_query q ON q.query_id = p.query_id
WHERE rs.last_execution_time >=
DATEADD(HOUR, -24, SYSUTCDATETIME())
GROUP BY q.query_id, p.plan_id
ORDER BY avg_ms DESC;
04 · SQL Server consulting for HA and DR
Always On engineered as a topology and proven by drills
An availability group that has never been failed over on purpose is an assumption. SQL Server consulting at MinervaDB designs the topology, then breaks it on a schedule to produce the numbers the business is relying on.
Figure 3. SQL Server consulting HA and DR reference topology: listener with read-only routing, synchronous secondary in the primary region, asynchronous secondary in the DR region, WSFC quorum with cloud witness, backups with an intact log chain, and the four numbers a drill produces.
Quorum designed for the failure you will have
SQL Server consulting for Always On starts at the cluster: node majority with a cloud or file-share witness, dynamic quorum, and cluster heartbeat and lease timeouts tuned for the actual network between sites. Quorum loss is rehearsed, because a DR site that cannot form quorum is not a DR site.
Synchronous where it counts
One synchronous secondary in the primary region for zero-data-loss automatic failover, asynchronous replicas across regions with the send queue alerted against the RPO budget, and readable secondaries carrying reporting through read-only routing so the primary’s CPU stays for transactions.
Backups with a log chain that restores
Full, differential and log backups with CHECKSUM and compression, copies to Azure Blob or S3 with immutability, and a full point-in-time restore timed quarterly on production-size data. That observed time is the RTO the runbook quotes, not the number on the slide.
05 · SQL Server consulting for versions and upgrades
SQL Server version lifecycle: what is out of support, and where to land
SQL Server 2014 left extended support in July 2024 and SQL Server 2016 on 14 July 2026; 2017 follows in October 2027. Unpatched instances carry unmitigated security exposure, and SQL Server consulting treats the upgrade as a first-class migration.
Figure 4. SQL Server version lifecycle as verified on 28 September 2026: support end dates by version, status today and the upgrade path MinervaDB uses for each.
| Upgrade method | Downtime | Best for | Rollback | Pre-flight evidence |
|---|---|---|---|---|
| AG rolling upgrade | Seconds at failover | Estates already on Always On; 2017 and 2019 to 2022 or 2025 | Fail back to the un-upgraded replica while it still exists | Query Store baseline captured, secondaries upgraded and synchronised first, compatibility level unchanged until regression review |
| Side-by-side migration | Minutes to an hour at cutover | Unsupported versions (2014, 2016) or hardware and OS refresh at the same time | Source instance retained; log shipping back where needed | Data Migration Assistant findings resolved, logins and jobs scripted, backups restored and verified on the new instance |
| In-place upgrade | Instance offline for the setup run | Single instances with a maintenance window and a verified restore | Restore of the pre-upgrade backup; the restore is timed first | Setup rules clean, disk space, SQL Agent and SSIS package inventory, third-party driver compatibility |
For SQL Server consulting targets today, SQL Server 2025 is the current release, with a native vector data type, native JSON type, optimized locking, regular expressions in T-SQL and REST endpoint support; it is the target for new builds and for modernisation where those features shorten application work. SQL Server 2022 remains the conservative target for lift-and-upgrade estates on its current cumulative-update cadence. Dates are from Microsoft’s lifecycle pages, verified today.
06 · SQL Server consulting for Azure SQL and cloud migration
Landing-zone selection by feature use and licence position, then a dual-run cutover
Azure SQL Database, Managed Instance, SQL Server on Azure VMs and Amazon RDS each remove a different set of responsibilities and impose a different set of constraints. SQL Server consulting picks the landing zone from what the estate actually uses.
Figure 5. The SQL Server consulting cloud migration decision map: assessment inputs, the four landing zones with their trade-offs, and the dual-run cutover engineering that keeps the source intact until sign-off.
Azure SQL Database
Single databases and elastic pools, Hyperscale for very large or bursty databases, no SQL Agent, CLR or cross-database queries. The lowest operations burden and the right home for isolated application databases.
Azure SQL Managed Instance
The SQL Server consulting default for lift-and-shift: near-full instance surface including Agent, CLR, linked servers and cross-database queries, with Managed Instance link for near-zero-downtime migration and hybrid DR. The default landing zone for instance-scoped estates.
SQL Server on Azure VMs or Amazon RDS
Full control, failover cluster instances and availability groups as on-premises, third-party constraints honoured; you still patch and operate. Azure Hybrid Benefit and reserved capacity modelled from measured utilisation.
Staged exit to PostgreSQL
Where per-core licensing no longer earns its keep, T-SQL and SQL Agent conversion is scored per object and the workload moves in waves with a dual-run CDC cutover. Data modernization →
07 · SQL Server consulting for security and compliance
Controls that produce evidence, not just settings
An assessor does not ask whether TDE is on; they ask for the key-rotation record, the access review and the audit export. SQL Server consulting at MinervaDB configures the controls and the evidence they generate together.
Figure 6. The SQL Server security and compliance control map: encryption, access control, auditing and detection, the evidence each control produces, and the framework mapping for GDPR, HIPAA, PCI DSS and SOC 2.
Encryption and access
SQL Server consulting for security configures TDE with keys in an EKM provider or Azure Key Vault and a rotation record, Always Encrypted with secure enclaves for data in use, TLS 1.2 or later enforced, backup encryption, least-privilege roles with no application in sysadmin, Entra ID or Windows authentication, row-level security and dynamic data masking for multi-tenant and non-production data.
Audit and evidence
SQL Server Audit to file or Azure Monitor retained for the compliance window, Extended Events on privileged actions, alerts on failed logins and permission changes, quarterly access reviews with role-membership snapshots, change records with approval and rollback, and DR drill reports, produced as by-products of operations rather than assembled before an audit.
08 · SQL Server performance audit
The fixed-scope entry point to SQL Server consulting at MinervaDB
Read-only, DMV and Query Store based, and delivered as findings your own team can verify. No change is made to production during the audit.
What is reviewed
SQL Server consulting audits cover instance and database configuration against workload and hardware (MAXDOP, cost threshold, memory, tempdb, auto-growth, compatibility levels), the wait profile, top statements and plan regressions, index usage and redundancy, statistics health, blocking and deadlocks, AG health and queues, backup and restore evidence, security posture and version currency.
What you receive
Findings ranked P0 to P2, each with the observation, the DMV or Query Store evidence, the recommended change with its rollback, and the metric it is expected to move; a prioritised remediation plan; and a versioned report your team keeps whether or not MinervaDB does the remediation. SQL Server performance audit →
Standing caveat: every recommendation on this page is tested on a non-production copy against production-representative data before it is applied to a production system, with a verified backup taken first and a disaster-recovery posture that has been exercised by a timed restore.
09 · SQL Server consulting engagement models
Audit, project, retainer, or 24×7 operations
Engage for one fixed-scope piece of SQL Server consulting or hand over the estate. Deliverables are versioned and execution-ready either way.
Performance audit
Fixed scope, read-only, findings within days of telemetry access. Sized by instance count and workload complexity.
Project delivery
An upgrade, an Always On build, an Azure migration or a security programme, delivered under the method above with rollback at every stage.
SQL Server consulting retainer
Senior SQL Server consulting hours each month for architecture, tuning and review, with the same engineers across the year.
24×7 remote DBA
Operations with defined SLAs (S1 15 min, S2 12 h, S3 24 h, S4 48 h), monitoring, patching, restore drills and a monthly SLO report. SQL Server support →
10 · FAQ
SQL Server consulting questions we are asked most
Short answers to what engineering leaders ask before the first call.
What does a typical SQL Server consulting engagement include?
A discovery call on workload, topology, pain points and licence position, then a scoped engagement: a performance audit, an Always On or DR review, an upgrade or Azure migration plan, a security programme, or ongoing remote DBA operations. Every engagement ends with written deliverables: findings with DMV and Query Store evidence, configuration changes with rollback, runbooks, and an action plan with the metric each item is expected to move.
Do you support both on-premises SQL Server and Azure SQL?
Yes. On-premises and VM-hosted SQL Server 2016 through 2025 including Always On availability groups and failover cluster instances, Azure SQL Database, Azure SQL Managed Instance, SQL Server on Azure VMs and Amazon RDS for SQL Server. The managed-versus-self-managed divergence for each service is documented before design starts.
Is 24×7 support available for production SQL Server environments?
Yes. A senior SQL Server engineer is on watch across APAC, EMEA and the Americas with S1 acknowledged within 15 minutes, S2 within 12 hours, S3 within 24 hours and S4 within 48 hours, with patching and cumulative-update management, backup verification, restore drills and a monthly SLO report.
Which SQL Server versions should we be on today?
SQL Server 2014 and 2016 are out of extended support and should be exited now; 2017 leaves extended support in October 2027 and 2019 is security-fix only. SQL Server 2022 is the conservative target and SQL Server 2025 is the current release with the vector type, native JSON and optimized locking. Every upgrade is run with a Query Store baseline, a staged compatibility-level change and a rehearsed rollback.
How is SQL Server consulting different from managed database services?
Consulting is scoped engineering work that returns findings, designs and changes with evidence; managed services or remote DBA is ongoing operation of the estate under SLAs. Most customers start with an audit or a project and move the platform into 24×7 remote DBA afterwards, with the same engineers accountable across both.
Can you reduce our SQL Server licence cost?
Often, and always from measurement: core counts against actual CPU utilisation, edition fit (Standard versus Enterprise features actually used), consolidation of under-used instances, Azure Hybrid Benefit and reserved capacity, and, where per-core licensing no longer earns its keep, a staged exit to PostgreSQL. We sell no licences, so the recommendation carries no commission.
How do you migrate to Azure SQL without downtime?
Assessment with Data Migration Assistant and Azure Migrate, landing-zone selection from feature use and licence position, then a dual-run cutover: Managed Instance link or a distributed availability group keeps the target current, parity is verified with row counts, checksums and Query Store replay, and traffic switches through the listener or connection string in a rehearsed window with the reverse path held open.
Is MinervaDB tied to Microsoft or any cloud?
No. We hold no licence resale or cloud partner incentives. SQL Server consulting recommendations are anchored in your wait statistics, Query Store data, drill results and invoice, including recommendations to stay where you are or to use a technology we do not sell services for.
Talk to a senior SQL Server consulting engineer
Bring a Query Store export, the wait statistics since the last restart, the last incident timeline and the current licence position to the first call. We will tell you which layer is the constraint, what moving it is worth, and what we would change first.