Different Types of Databases

A. Caching Solutions

Requirements:

1. Fast
2. Key-Value Pairing
3. Data persistence is not required (data loss is acceptable)

Best to use: Redis, MemCache, etcd

B. File Storage (Images, Videos)

Requirements:

1. Ability to store large binary objects (Blob storage)
2. No need for relational queries or joins on stored data

Best to use: Amazon S3

Note: Pair File-Storage with CDN for faster ascess

C. Text Searching

Requirements:

1. Keyword-based search
2. Autocomplete / prefix matching (e.g., ap → apple)
3. Fuzzy search (handling typos like apjle → apple)

Best to use: Search-Engines like Elastic Search or Solr (Apache Lucene)

D. Metrics/ Time series data

Requirements:

1. Optimized for time-based data
2. Data written sequentially (append-only pattern)
3. Queries typically involve time ranges and aggregations
4. High write throughput & Time-based partitioning

Best to use: InfluxDb, TimeScaleDb

E. Big Data Analysis (Data Warehouse)

Requirements:

1. Ability to store and process very large datasets (TBs–PBs), Fast Read Throughput.
2. Optimized for analytical queries and aggregations
3. Used mainly for offline reporting and business intelligence
4. Supports OLAP-style queries over large datasets

Typical OLAP Questions
Examples companies ask:
"What was total revenue per country last year?"
"Which product category grew fastest?"
"What is average user session duration?"
"How many orders per hour?"

Best to use: Hadoop, Redshift, BigQuery, Snowflake

F. Sql

1. Data is structured and follows a predefined schema.
2. Requires strong transaction guarantees (ACID).
3. Suitable for systems where data consistency is critical.
4. Example: Banking Payment Systems

Best to use: Postgres, mySql

G. Document Db (NoSql)

1. Data is semi-structured (JSON-like documents).
2. Schema is flexible and can evolve easily.
3. Documents may contain many attributes/fields that vary between records.
4. Applications need to query and index these attributes efficiently.
5. Example: User profiles, Product catalogs

Best to use: MonogoDb (Supports ACID), CouchBase

H. Columar Database (NoSql)

1. Designed for very large datasets distributed across many machines.
2. Data is typically append-heavy and continuously growing.
3. Optimized for high write throughput and horizontal scaling.
4. Queries often access specific columns across large datasets.

Best to use: CassandraDb, Hbase

Summary

Type	Best For	Read Throughput	Write Throughput	ACID / BASE	Examples
Cache	Fast temporary data, frequently accessed items	Very High	High	BASE	Redis, Memcached
Object Storage	Large files (images, videos, backups)	High	Medium	BASE	Amazon S3, Google Cloud Storage
Search Engine	Full-text search, autocomplete, fuzzy search	High	Medium	BASE	Elasticsearch, Solr
Time Series DB	Metrics, monitoring data, IoT telemetry	High	Very High (append-heavy)	BASE	InfluxDB, Prometheus
SQL (Relational DB)	Structured data with strong consistency and transactions	High	Medium	ACID	PostgreSQL, MySQL
Document DB	Flexible schema, JSON-like data, evolving structures	High	High	Mostly BASE (some ACID support)	MongoDB, Couchbase
Wide Column DB	Massive distributed datasets, high write scalability	Medium	Very High	BASE	Cassandra, HBase
Data Warehouse	Large-scale analytics and reporting (OLAP)	Very High (large scans)	Low–Medium	BASE	BigQuery, Redshift, Snowflake