Distributed Computing: Distributed System Checkpoints and it’s type-checkpoint levels its tools.<\/p>\n
[fvplayer id=”59″]<\/p>\n
Checkpointing is a fault-tolerance mechanism<\/strong> in distributed computing that periodically saves the system state. If a failure occurs, the system can restart from the last checkpoint<\/strong> instead of starting from scratch.<\/p>\n Key Idea:<\/strong> Definition:<\/strong> All nodes in the system synchronize and save their states together.<\/p>\n Ensures consistency<\/strong> (no orphan or lost messages). Example:<\/strong><\/p>\n Definition:<\/strong> Each process takes checkpoints independently<\/strong> without synchronization.<\/p>\n Faster<\/strong>, no coordination required. Example:<\/strong><\/p>\n Definition:<\/strong> A hybrid approach where checkpoints are triggered<\/strong> based on message passing.<\/p>\n Prevents inconsistent states.<\/strong> Example:<\/strong><\/p>\n Definition:<\/strong> Checkpoints are managed at the software level<\/strong> rather than system level.<\/p>\n \u00a0Allows customized checkpointing<\/strong> in applications. Example:<\/strong><\/p>\n Process-Level Checkpointing:<\/strong> Saves state of individual processes. 1. DMTCP (Distributed MultiThreaded Checkpointing)<\/strong><\/p>\n 2. CRIU (Checkpoint\/Restore in Userspace)<\/strong><\/p>\n 3. BLCR (Berkeley Lab Checkpoint\/Restart)<\/strong><\/p>\n 4. Hadoop Checkpointing<\/strong><\/p>\n Checkpointing reduces system failures’ impact<\/strong> by allowing recovery from saved states. Different types & tools<\/strong> are used based on the system requirements<\/strong> (speed, reliability, and overhead).<\/p>\n Would you like code examples or real-world use cases<\/strong>?<\/p>\n Distributed Computing<\/strong> involves multiple computer systems working together to achieve a common goal. One key challenge in such systems is ensuring fault tolerance<\/strong>, which is where checkpoints<\/strong> come in.<\/p>\n A checkpoint<\/strong> is a saved state of a process or the entire system at a specific point in time. If a failure occurs, the system can roll back<\/strong> to the last checkpoint rather than starting over.<\/p>\n Fault Tolerance<\/strong><\/p>\n<\/li>\n Failure Recovery<\/strong><\/p>\n<\/li>\n Performance Optimization<\/strong><\/p>\n<\/li>\n Minimizing Downtime<\/strong><\/p>\n<\/li>\n<\/ul>\n Each process independently saves its state.<\/p>\n<\/li>\n Simple but inconsistent<\/strong>: may lead to domino effect<\/em> (rollback cascades).<\/p>\n<\/li>\n<\/ul>\n A set of local checkpoints, one per process, such that the combination represents a consistent global state<\/strong>.<\/p>\n<\/li>\n Used in coordinated checkpointing.<\/p>\n<\/li>\n<\/ul>\n
\u00a0Saves system state at intervals.
\u00a0Reduces computation loss during failures.
\u00a0Speeds up recovery in distributed systems.<\/p>\n\u00a0Types of Checkpoints in Distributed Systems<\/strong><\/h3>\n
Coordinated Checkpointing<\/strong><\/h3>\n
\u00a0Used in global snapshots<\/strong>.
Slower<\/strong> due to coordination overhead.<\/p>\n\n
\u00a0Uncoordinated Checkpointing<\/strong><\/h3>\n
Risk of cascading rollbacks (domino effect).<\/strong><\/p>\n\n
\u00a0Communication-Induced Checkpointing<\/strong><\/h3>\n
Avoids the domino effect<\/strong> from uncoordinated checkpointing.
Higher overhead due to message tracking.<\/strong><\/p>\n\n
\u00a0Application-Level Checkpointing<\/strong><\/h3>\n
\u00a0Efficient for high-performance computing (HPC)<\/strong>.
Requires developer implementation.<\/strong><\/p>\n\n
\u00a0Checkpoint Levels in Distributed Systems<\/strong><\/h3>\n
System-Level Checkpointing:<\/strong> Saves the entire OS state.
Application-Level Checkpointing:<\/strong> Saves the state at the application level.<\/p>\n\u00a0Tools for Checkpointing in Distributed Systems<\/strong><\/h3>\n
\n
\n
\n
\n
\u00a0Conclusion<\/strong><\/h3>\n
\n\ud83e\udde9 What is a Checkpoint in Distributed Systems?<\/strong><\/h2>\n
\n\ud83d\udee0\ufe0f Purpose of Checkpointing<\/strong><\/h2>\n
\n
\n\ud83d\udcca Types of Checkpoints in Distributed Systems<\/strong><\/h2>\n
1. Local Checkpoints<\/strong><\/h3>\n
\n
2. Global Checkpoints<\/strong><\/h3>\n
\n
\n\ud83e\udded Checkpointing Levels<\/strong><\/h2>\n