Promotion Black Fridayjusqu'à 60% de réduction
  • 60% de réduction sur votre première année d'abonnement annuel
  • 45% de réduction sur les licences à vie
Se termine dans
Profiter de l'offre
Catégorie

Infrastructure & Scaling

Servers, performance, deployments and backups for growing platforms.

10 articles

The Result Was Fast. Its Queue Was Not — NovaCommerce 4
Redis & cachingQueues & workersObservability

The Result Was Fast. Its Queue Was Not — NovaCommerce 4

A 0.34-second result job waited behind slow AI work. Separate lanes helped, but durable submissions, bounded retries and worker recovery still mattered.

Anichur Rahamanil y a 1 mois5min
The Database Moved. One Old Client Did Not — NovaCommerce 3
DatabasesVector search (pgvector)Backups

The Database Moved. One Old Client Did Not — NovaCommerce 3

A 26-minute database cutover looked complete until a legacy panel wrote to the old copy. The lesson is writer control, reconciliation and honest recovery.

Anichur Rahamanil y a 1 mois5min
Ready Before the Exam: Scaling and Replacing Web Nodes — NovaCommerce 2
Load balancingAutoscalingDevOps

Ready Before the Exam: Scaling and Replacing Web Nodes — NovaCommerce 2

Two application pools helped a timed-exam platform grow. The harder work was proving readiness, draining old nodes and keeping one version of the service.

Anichur Rahamanil y a 1 mois5min
A Smaller Cloud Bill—or a Bigger Operations Job? Price the Whole Move
Cloud costsSelf-hostingData sovereignty

A Smaller Cloud Bill—or a Bigger Operations Job? Price the Whole Move

Compare the service you need, not a cloud invoice against a bare server. Include people, recovery and exit costs before choosing managed hosting, repatriation or a hybrid.

Anichur Rahamanil y a 1 mois6min
Checkout Stuck, CPU Calm: Observability That Finds the Unfinished Work
ObservabilityDevOpsPerformance

Checkout Stuck, CPU Calm: Observability That Finds the Unfinished Work

A green server graph can hide delayed orders. Connect real business outcomes to metrics, protected logs and traces, then give each alert an action and an owner.

Anichur Rahamanil y a 1 mois6min
Protect the Checkout Data Path: Pools, Replicas, Cache and Vector Search
DatabasesRedis & cachingVector search (pgvector)

Protect the Checkout Data Path: Pools, Replicas, Cache and Vector Search

A connection limit is not a capacity plan. Budget every pool, define which reads may be stale and keep disposable cache separate from business obligations.

Anichur Rahamanil y a 1 mois8min
Queues That Finish the Work: Capacity, Ownership and Safe Recovery
Queues & workersAutoscalingRedis & caching

Queues That Finish the Work: Capacity, Ownership and Safe Recovery

A fast response can hide unfinished business work. Size the worker pool for arrivals and backlog, preserve the accepted record and make retries accountable.

Anichur Rahamanil y a 1 mois8min
Ransomware-Ready Backups: Prove the Business Can Recover
BackupsCybersecuritySelf-hosting

Ransomware-Ready Backups: Prove the Business Can Recover

Use 3-2-1-1-0 without counting a fourth copy by accident. Test protected versions, recovery keys, actual RPO/RTO and a clean business restart.

Anichur Rahamanil y a 1 mois8min
Ready Before the Spike: A Capacity Plan for the Whole Request Path
AutoscalingLoad balancingServer scaling

Ready Before the Spike: A Capacity Plan for the Whole Request Path

Adding an app instance is not the same as serving another checkout. Budget readiness time, test the constrained tier and keep spare capacity for failure.

Anichur Rahamanil y a 2 mois9min
Scaling Self-Hosted Commerce: From Daily Orders to a Tested Capacity Plan
Server scalingDevOpsPerformance

Scaling Self-Hosted Commerce: From Daily Orders to a Tested Capacity Plan

100,000 orders a day is a business goal, not a server specification. Translate traffic into peak work, remove the measured bottleneck and test failure capacity.

Anichur Rahamanil y a 4 mois8min