Price: 0
Number of applications: 2
12.12.25 (inclusive)
By agreement
Idea
ICT tasks
Food industry
Intelligent control systems
Software/ IS
In many PostgreSQL products, replication is configured "manually" for each project, which leads to: • disparate configurations and scaling errors; • difficulties with updating PostgreSQL versions; • Lack of a single standard for CDC and integrations with external systems; • difficulties in diagnosing replication problems and standby server backlog. As the load and volume of data increases, this leads to lower productivity, risks of data loss, and the inability to quickly build storefronts/analytics. A unified, reproducible solution is needed that: • is deployed via Docker Compose, • Uses Debezium for reliable CDC, • Easily portable between environments (dev/stage/prod), • it is documented and understandable to other teams/partners.
• Reducing the deployment time of the PostgreSQL replication loop from several days to several hours due to ready-made docker templates. • Improve replication reliability and reduce the number of incidents related to data lagging or out of sync. • The ability to build real-time analytics and integrations through CDC (Debezium/Kafka) without complex modifications to the main database. • Reducing the load on the main cluster due to reporting and heavy requests for replicas. • Create an in-house replication configuration standard that can be scaled to new projects and franchise locations.
Nurlan Rakhimzhan
Purpose and description of task (project)
The goal of the task is to develop, test and document a single standard PostgreSQL replication contour for high-load SaaS systems (using the example of the Alarify ERP platform), using Docker Compose, logical replication and the Debezium/Kafka stack. The project plans to: • describe typical replication scenarios (master → read-replica, master → reporting DB, master → CDC bus); • prepare a docker-compose environment with the configuration of PostgreSQL (primary/standby), Debezium, Kafka and auxiliary services; • Implement a Debezium-based change data capture (CDC) mechanism to transfer changes to analytical and integration circuits; • set up monitoring and alerting for replication delays and errors (Prometheus/Grafana/logging); • prepare a basic set of best practices for configuring PostgreSQL parameters for stable logical replication in production. The result will be a reusable solution (docker templates, configurations, documentation) that can be used in any ERP/CRM/SaaS systems with microservice architecture.
Whatsapp
