Platforms / 12
Connect SQL data. Preserve application consistency.
Reporting, portals or business applications need data from another database. We plan a connection that accounts for access rights, consistent data states and the rules of the applications involved.
The starting point
Copying tables can bypass business rules.
An overnight export extracts customers and orders at different times. Both tables change during extraction. Comparing row counts alone does not prove the target dataset is consistent for the business.
We assess database type and version, supported views, stable keys and permitted write paths. Direct changes to another application's database can affect validation, triggers or vendor support.
What DataSource delivers
Extraction. Staging and mapping. Target integration.
Extraction
Define required fields and a consistent read state. Select views, snapshots or change data capture according to actual capabilities.
Staging and mapping
Validate data first and map business IDs. Check types, time zones, null values and relationships before import.
Target integration
Write through approved APIs, imports or explicitly supported database procedures. Document transfer progress and reconciliation rules.
SELECT id, updated_at, customer_number
FROM integration_customers
WHERE updated_at > :last_time
OR (updated_at = :last_time AND id > :last_id)
ORDER BY updated_at, id;
-- Cursor erst nach bestätigter Übernahme fortschreiben.Technical scenario / Illustrative example
Read changes without losing concurrent updates.
A composite cursor helps when change timestamps are identical. This SQL is a design example, not a universal CDC replacement: reliable timestamps, consistent reads and deletion notifications remain necessary.
Example data and proposed rules. The actual interfaces and business rules are assessed for your project.
Error handling, security and operations
Permissions: validate and operate.
Permissions
Restrict read-only accounts to approved objects. Agree on network access, encrypted connections and parameterized queries.
CDC limitations
CDC depends on the database. PostgreSQL logical replication, for example, does not automatically transfer schema changes; DDL and sequences require a separate design.
Load and operations
Agree on batch sizes, indexes and runtimes with database operations. Long transactions and change-data retention can put load on the source system.
Technical questions
Details that matter before implementation.
SQL · Staging · Snapshots · Delta · CDC · Transactions
Is an updated_at field sufficient?
Only when changes are captured reliably. Backdated changes, deletions and concurrent transactions require additional rules or a different change-tracking mechanism.
Can different databases be connected?
Yes, with a suitable data model and supported access. SQL dialects, decimal values, collations and time types need to be addressed explicitly.
When is an API preferable?
When the application encapsulates business rules and permissions through its API. Database access is then mainly appropriate for approved extraction.
For technical clarification
Start with your actual workflow.
Identify database type and version, permitted views, data volume, change frequency and existing vendor requirements. Schema extracts and anonymized relationships are sufficient for initial clarification.
In our initial discussion, we clarify available access, data ownership and a narrowly scoped first data flow. We then agree on scope, responsibilities and commercial terms.
Anonymized examples are enough to get started. Please do not send passwords, tokens or personal data from production systems in your initial inquiry.
Related challenges
Your next connection in context.
Synchronize data between systems
A customer is updated in the CRM and in the ERP. Which version is authoritative? We design synchronization around business ownership and stable identities so that recurring transfers do not create silent data conflicts.
Legacy systems & middleware
The legacy system serves its purpose but does not offer the same interfaces as a new application. We examine existing access options and develop narrowly scoped middleware to bridge the differences in a controlled way.
Automate reporting, PDFs & documents
A report is assembled from several exports, a template is filled in manually and distributed as a PDF. We connect data preparation and document generation in a workflow that makes origin, version and approval traceable.
Technical references
Official documentation, reviewed on October 6, 2026. Suitability depends on the specific installation and version.