Microsoft SQL Server development in Kolkata

+91 94329 43298 WhatsApp

Microsoft SQL Server development

The database a great many established business systems in this region already run on — and where most of their performance problems live.

Relational database

Chosen per project

Maintainable handover

Where a business has been running packaged or bespoke software for a decade, the odds are good that SQL Server is underneath it. That makes it less a technology choice than a fact to work with — and usually the cheapest place to improve a slow system.

The pattern we see repeatedly is a schema that grew by addition, indexes created for problems long gone, and a handful of queries responsible for most of the load. Finding those is a day of work; rewriting the application is months.

What we build with SQL Server

Existing .NET and ERP systems built on SQL Server

Reporting that has begun to interfere with daily operations

Integrating a legacy database with a new application or API

Migration to Azure SQL or a managed instance

Where it fits — and where it does not

Good fit when

Transactional business systems already on the platform

Organisations standardised on Microsoft tooling

Complex reporting needing window functions and stored procedures

Consider something else when

Greenfield projects with no Microsoft dependency — PostgreSQL avoids the licence

Document-shaped data with no fixed schema

Performance work

How a slow system gets diagnosed

In that order. Reversing it is how people end up buying hardware that does not help.

Find the top queries

Query Store ranks by total load, not by complaint.

Read the plan

Scans, spills and bad estimates show up here.

Fix indexes or SQL

Usually one index or one rewrite carries most of the gain.

Re-measure

Confirm on real load before touching anything else.

How we work with SQL Server

Measure before changing

Query Store and execution plans identify the small number of statements causing most of the load, before anything is rewritten.

Indexes with intent

Missing indexes added, unused and duplicate ones removed — the second half is skipped surprisingly often, and it costs on every write.

Set-based rewrites

Row-by-row cursors replaced with set operations where that is what is actually slow.

Backups proved

Full and log backups on a schedule, with a restore actually performed rather than assumed.

Our typical SQL Server setup

Typical SQL Server stack choices and what we use for each
ConcernWhat we use
VersionsSQL Server 2016 onwards, Azure SQL, Managed Instance
DiagnosisQuery Store, execution plans, wait statistics
SchemaNormalised transactional design, with reporting views where useful
ProgrammabilityStored procedures and functions, kept in source control
OperationsBackup, restore testing and index maintenance jobs

Frequently asked questions

Usually. It is typically reporting queries scanning large tables while transactions are running. Options range from an index and a rewrite to moving reporting onto a replica — we would measure first and recommend the cheapest effective one.

Only if the migration cost is genuinely repaid. For an existing application with stored procedures and .NET data access, that is a substantial rewrite, and moving to a managed SQL Server tier is often the better economics.

Yes, and that is most of this work. It begins with reading the schema and the top queries rather than proposing changes.

Topics people search for

Services built with SQL Server

Tell us what you are trying to build

Describe the problem in plain language and we will tell you what it would take to solve it — the approach, the moving parts and the sensible order to build them in. No obligation either way.

WhatsApp
Call now Enquire