How to Build a Startup Backend Technology Stack from Scratch

For startup founders, building a backend stack involves selecting languages, components, processes, and systems that together form the foundation of all server‑side services.

Four Layer Overview

Language: C++, Java, Go, PHP, Python, Ruby, etc.

Components: MQ, databases, caching, etc.

Process: development, release, monitoring, code standards.

System: tools that enforce the above processes (e.g., release platforms, code repositories).

Component Selection

1. Project/Bug Management

Open‑source options include Redmine (Ruby), Phabricator (PHP), Jira (Java) and Wukong CRM for B2B customer‑centric workflows.

2. DNS

Domestic choices are Alibaba Wanwang and Tencent DNSPod; for global reach, Amazon Route 53 is recommended.

3. Load Balancing (LB)

Cloud providers offer LB services (Alibaba SLB, Tencent CLB, AWS ELB); self‑hosted environments typically use LVS + Nginx.

4. CDN

In China, NetEase (Wangsu), Tencent and Alibaba dominate; internationally, Amazon CloudFront and Akamai are common.

5. RPC Frameworks

Cross‑language options: Thrift, gRPC, Hessian, Hprose. Service‑governance focused frameworks: Dubbo, DubboX, Motan, rpcx.

6. Service Discovery

Popular registries: etcd, Consul, Apache ZooKeeper; custom implementations or Redis are also possible.

7. Relational Databases

Traditional: MySQL/MariaDB; NewSQL examples: CockroachDB, TiDB, Google Spanner.

8. NoSQL

Key‑value (Redis, Memcached), columnar (HBase, Cassandra), document (MongoDB, CouchDB), graph (Neo4j).

9. Message Middleware

Used for async processing, system decoupling, and traffic shaping; selection can be open‑source or custom (e.g., RabbitMQ, Kafka).

10. Code Management

Git is essential; GitLab (self‑hosted) combined with Gerrit provides robust code review.

11. Continuous Integration

Options include Jenkins, TeamCity, Strider, GitLab CI, Travis, GoCD.

12. Logging

ELK stack (Elasticsearch, Logstash, Kibana, Filebeat) with Nginx reverse proxy for security.

13. Real‑Time Analytics

Typical pipeline: Flume → Kafka → Storm → MySQL.

14. Monitoring

Prometheus + Grafana is preferred for modern stacks; alternatives include Zabbix and Open‑Falcon.

15. Configuration Management

ZooKeeper/etcd with UI or automation tools (Puppet, Ansible) for dynamic config.

16. Release/Deployment

Workflow: code → artifact → deployment → production; tools such as Jenkins, GitLab CI, Walle, Piplin can be combined.

17. Jump Server

Jumpserver (open‑source) provides role‑based access, auditing, and command recording.

18. Machine Management

Ansible is recommended for its agent‑less design; alternatives include Puppet, Chef, SaltStack.

Startup‑Specific Guidance

1. Choose the Right Language

Select a language familiar to the team, modern with built‑in memory management, with strong open‑source ecosystem and hiring pool.

2. Choose Components and Cloud Provider

Prefer reliable cloud vendors and mature open‑source components that have proven adoption in large internet companies.

3. Define Processes and Standards

Establish development, release, operations, database, alerting, and reporting procedures.

4. Build or Adopt Supporting Systems

Use open‑source tools that match the chosen language and component stack to solidify processes.

5. Consider Long‑Term Implications

Evaluate cost, time, scalability, and future migration needs when locking in a technology stack.