Mugdha

Sep 15, 2023

Caches

A cache in a distributed system is similar to a cache in a single system but is extended to a network of interconnected computers or nodes. In distributed systems, a cache is a mechanism that stores copies of frequently accessed data or computation results to reduce the need to get that data or recompute the results from their original source, which might be remote or relatively slow to access.

Aspects of cache:

Data replication: Caches in distributed systems replicate data or computation across multiple nodes or locations to improve access times for all nodes that need the same data.

Network efficiency: Because caches reduce the need for frequent data transfers across a network, caches thereby help improve network efficiency and reduce latency.

Load Balancing: Caches can distribute data or computation load across nodes ensuring that no single node becomes a bottleneck for requests.

Fault Tolerance: Caches can be used to provide fault tolerance by allowing nodes to continue serving data even if the primary source is temporarily unavailable.

Consistency and Invalidation: Data consistency is crucial for distributed systems. Caches implement strategies for cache invalidation or consistency protocols to ensure that cache data remains accurate and up to date.

Cache Coherency: In distributed systems, multiple nodes may have their own copies of cached data. Cache coherency protocols ensure that these copies remain consistent and synchronized when changes are made.

Cache types: Distributed systems can employ various types of caches including in-memory caches, distributed caches(eg. Redis, Memcached), and Content Delivery Networks (CDNs) for caching web content.

Scalability: Caches are important for Scaling distributed systems. So, as the system grows, caching can help manage the increased load on backend resources.

Types of caches and examples:

In-Memory cache: Redis, Memcached, EHCache, Guava Cache.

Distributed Cache: Redis, Memcached, Hazelcast, Apache Ignite.

CDN: Cloudflare, Akamai, Amazon Cloudfront, Fastly.

Web Browser Cache: Browser-specific caches like Chrome Cache and Firefox Cache.

Proxy Cache: Nginx, Squid, Varnish.

Database Query Cache: MySQL query cache, PostgreSQL query cache extensions.

Object Caching: .NET Memory Cache, Java Caffeine Cache, PHP APCu.

API Response Cache: AWS API gateway caching, Nginx caching for API endpoints.

Frontend Component Cache: Angular Universal Server-Side Rendering Cache, React Server Components Cache.

Message Broker Cache: Rabbit MQ message Cache, Apache Kafka in-memory message buffer.

Metadata Cache: Amazon S3 Object Metadata Cache, Schema Metadata Cache in relational databases.

Session State Cache: ASP. NET session state cache, Redis for session state.

Service Discovery Cache: Consul, Eureka, Etcd.

Gateway Cache: Spring Cloud Gateway Cache, Kong Gateway Cache.

Session Cache: Tomcat session manager, ASP.NET session state cache.

HTTP Caching: Browser Cache(Chrome Cache, Firefox Cache)

Message Queue Cache: Rabbit MQ, Apache Kafka, Amazon SQS.

User Authentication Cache: OAuth 2.0 token caches, JWT token caches.

Template Caching: Smarty Template Caching, Django Template Caching.

Dependency Resolution Cache: Maven Central Repository Cache, npm package cache.

GPU Memory Cache: NVIDIA CUDA GPU memory cache.

Content Cache in Streaming Services: CDN Caches for Video Streaming, like Netflix's Open Connect.

Machine Learning Model Cache: MLFlow Model Cache, Tensorflow Serving Caching.