35+ YEARS IN MARKET
Continuously operating since 1990. 25,000+ consulting hours delivered and 596+ critical production incidents resolved. Pattern recognition only decades of real operation can build.
REMOTE SQL SERVER DBA — SINCE 1990
Microsoft sells it as an easy product. In production, it's Always On with a replica out of sync, tempdb locking, Query Store down and MAXDOP on default. We handle it 24/7 — in a physical facility, with SLA.
SQL Server environments managed by HTI
EVERYONE'S PROBLEMS
Tempdb locking at 2 PM every day
PAGELATCH_UP on GAM/SGAM/PFS, contention on allocation pages. Adding files isn't enough — needs trace flag 1118/1117, partitioning and workload sizing. Who knows that?
Always On with a suspended replica
Log send queue growing, redo queue stuck, automatic failover that doesn't happen. Without alerts, no one sees it. By then, RPO is gone. AG monitoring is more than CHECKDB.
Query Store off in production
Modern SQL Server's most useful feature is disabled in most environments. Without it, plan regression becomes mystery. With it, plan forcing fixes it in 5 minutes.
HTI runs the watch your team doesn't
We collect wait-stats baseline, monitor AG sync, keep Query Store active and versioned, open Microsoft tickets when needed. You only hear about it when you gained time.
MONTHLY SCOPE
Monthly contract with contractual SLA. Not hourly billing, not on-call freelancer. Permanent team operating your SQL Server as if it were ours.
Continuous collection from sys.dm_os_wait_stats with historical baseline. CXPACKET, PAGEIOLATCH_SH, WRITELOG, RESOURCE_SEMAPHORE — all categorized, all trended.
Sync state monitoring, log send/redo queue, automatic failover tested monthly. Listener health, read-only routing, backup on secondary.
Query Store enabled, controlled plan forcing, identification of top regressed queries. Plan tuning without changing code when possible.
Tempdb sizing and partitioning by workload, min/max memory tuning, PAGELATCH on allocation pages identification, correct trace flags for your version.
Full + diff + log backup with controlled VLF. Restore tested in staging monthly. Offsite copy and documented PITR strategy.
Audit of Enterprise vs Standard usage, features that justify the upgrade — or that don't. Defense against Microsoft Software Asset Management.
NO FREELANCERS. NO CALL CENTER.
The DBAs monitoring your SQL Server tomorrow are the same as today. Knowledge accumulates. Shift change uses formal handover, not prayer.
We don't sell hours. We sell sustainment. In December 2024, our SQL Server clients had zero out-of-SLA incidents.
PROOF IN PRODUCTION
CAPPTA — PAYMENTS
Black Friday + Christmas
SQL Server 2019 Always On AG on critical payments. Black Friday and Christmas with no maintenance window, no transaction loss, with 10x normal load peak.
MAJOR RETAILER
Tempdb redesign
PAGELATCH_UP eliminated with correct tempdb partitioning, TF1118 and allocation page tuning. Daily 2 PM slowdown gone.
HEALTHCARE
Multi-region DR
Always On AG across two regions + log shipping to a third, with documented PITR. LGPD compliance met without increasing license.
ONBOARDING
Environment discovery
Inventory of instances, editions, existing AGs, SQL Agent jobs, linked servers and replications. Criticality mapping by database.
Wait stats baseline
7-day collection of wait stats, perfmon counters and Query Store. Immediate identification of top 5 latent issues.
Documentation and runbook
Per-instance runbook with topology, dependencies, contacts, escalation. All versioned in HTI internal git.
Go-live 24/7 NOC
Calibrated alerts, tested escalation, first monthly report scheduled. You only get a call when you need to decide, not to inform.
FAQ
We do. Express has RAM and DB size limits, but the operation still deserves monitoring. Web Edition is common in hosting; we manage with documented restrictions. In general we recommend Standard for any production environment.
Yes, since the 2017 release. Practices change little (same engine); what changes is OS tooling — we use perf, top, iostat instead of Performance Monitor.
If the client has Premier, we open tickets on their behalf and follow up. If not, we have direct contacts with Microsoft technical escalation for cases that need product support. We never push a ticket and disappear.
We do when contractually agreed. The default is to point out the problem query with plan and impact proof; many clients prefer that the internal dev applies the change. In environments without available devs, we refactor the SP with documented tests.
We do it. Includes DMA compatibility analysis, workload capture, choice between Managed Instance vs SQL on VM vs RDS, and cutover with minimal window. Not a separate service — whoever operates your base also migrates it.
20 minutes with a Senior SQL Server DBA. Tell us your pain (tempdb, AG, seasonal slowness) and we'll tell you what we can solve in 30 days.
Versions and editions covered
WHY HTI
Sustaining critical environments in Brazil since 1990. We're not a generic consultancy that learned databases — we are The Database Company.
Continuously operating since 1990. 25,000+ consulting hours delivered and 596+ critical production incidents resolved. Pattern recognition only decades of real operation can build.
The only MySQL Authorized Education Center (AMEC) of Oracle University in Brazil — achieved twice. Over 1,000 DBAs trained in official HTI courses.
DBAs operate from a restricted-access room at PIT São José dos Campos: 24h CCTV, biometric access, dedicated rack. No remote freelancers, no improvised workstations. Compliance that goes beyond internal policy.
Behavioral monitoring developed in-house. Detects anomalous ingestion patterns and access to protected objects — no third-party tool dependency.