---
title: "MapReduce with GFS and HDFS"
subject: "Information System"
chapter: 8
key: "semester_8/information_system:mapreduce-gfs-hdfs"
source: https://www.bctengineer.com/q/semester_8/information_system/mapreduce-gfs-hdfs
site: BCT Engineering
---

# MapReduce with GFS and HDFS

Information System · Chapter 8: Scalable and Emerging Information System techniques

- Asked in: 2073 Bhadra, 5+5 marks
- Video: [Google File System - Paper that inspired Hadoop](https://www.youtube.com/watch?v=eRgFNW4QFDc) — Defog Tech
- Answer checked by Claude Opus 5.5 (high effort) on 2026-10-14
- Page: https://www.bctengineer.com/q/semester_8/information_system/mapreduce-gfs-hdfs
- Whole chapter: https://www.bctengineer.com/md/semester_8/information_system/chapter-8

## Question

What makes the Map Reduce as the de facto standard of scalable filing system? How closely Map Reduce is associated with GFS and HDFS? Explain in brief.

## Short answer

MapReduce's simple map/reduce model, data locality, fault tolerance, and scalability make it standard. It runs on GFS (Google) or HDFS (Hadoop) which store large files as replicated blocks and provide locations so MapReduce schedules tasks near data.

## Answer

#### Why MapReduce is the de facto standard for scalable processing of big data
MapReduce is strictly a processing model, not a file system; it became the standard for scalable big data processing on top of distributed file systems for these reasons:
- **Simple model:** the programmer writes only map and reduce; parallelism, distribution and failure handling are automatic.
- **Scalability:** runs on hundreds or thousands of cheap commodity machines; adding nodes increases speed almost linearly.
- **Data locality:** moves computation to the node holding the data, saving network traffic.
- **Fault tolerance:** failed tasks are re-executed; data is replicated.
- **Wide applicability:** indexing, log analysis, sorting, graph and machine learning jobs.
- **Open-source implementation:** Hadoop made it available to everyone, and many tools (Hive, Pig) are built on it.
- **Proven by Google** on web-scale data, at low cost.

#### Association with GFS and HDFS
- **GFS (Google File System)** is Google's distributed file system. It stores huge files as 64 MB chunks, replicated (3 copies) on chunk servers, with a single master holding metadata. MapReduce at Google reads its input from and writes its output to GFS.
- **HDFS (Hadoop Distributed File System)** is the open-source equivalent. It uses a NameNode (metadata) and DataNodes (blocks of 128 MB, replicated 3 times). Hadoop MapReduce uses HDFS in the same way.
- **Close link:** the file system provides large, replicated, sequential-access storage and tells MapReduce where blocks are, so that the scheduler places map tasks near the data. MapReduce provides the processing layer on top.

```text
 MapReduce (processing)
        |
 GFS / HDFS (storage, replication, locality)
        |
 Commodity machines
```

## Check yourself

The answer's key points as multiple-choice questions, with the correct option and why.

1. Which feature lets the programmer ignore parallelism, distribution and failure handling?
   A. Its simple model where only map and reduce are written
   B. Its use of expensive specialised hardware
   C. Its requirement for a fixed schema
   D. Its random-access storage
   Answer: A. The framework automates parallelism, distribution and failure handling, making the model simple. (Key point: Why MapReduce is the standard)

2. How does GFS store a huge file?
   A. As one file on a single server
   B. As 64 MB chunks, replicated 3 times on chunk servers, with a master holding metadata
   C. As 128 MB blocks managed by a NameNode
   D. As rows in a relational table
   Answer: B. GFS uses 64 MB chunks, 3 replicas and a single master for metadata. 128 MB blocks and NameNode are HDFS. (Key point: GFS design)

3. In HDFS, what do the NameNode and DataNodes hold?
   A. NameNode holds blocks; DataNodes hold metadata
   B. NameNode holds metadata; DataNodes hold replicated blocks
   C. Both hold only metadata
   D. Both hold only compute tasks
   Answer: B. The NameNode keeps metadata; DataNodes store the 128 MB blocks, replicated 3 times. (Key point: HDFS components)

4. How does the file system help MapReduce scheduling?
   A. It executes the reduce function
   B. It tells MapReduce where blocks are so map tasks are placed near the data
   C. It sorts the final output
   D. It compiles the map code
   Answer: B. Knowing block locations lets the scheduler apply data locality. (Key point: File system and MapReduce link)

## Related in Information System

- [GFS vs HDFS architecture](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Agfs-vs-hdfs) — Information System, chapter 8
- [MapReduce vs Hadoop; IR process](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Amapreduce-vs-hadoop-ir) — Information System, chapter 8
- [MapReduce working with example](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Amapreduce-working-big-data) — Information System, chapter 8
- [MapReduce vs Hadoop](https://www.bctengineer.com/question.md?key=semester_8%2Finformation_system%3Amapreduce-vs-hadoop) — Information System, chapter 8

## Related in other subjects

- [K-map Σ(1,2,3,8,10,13)+d, standard SOP](https://www.bctengineer.com/question.md?key=semester_3%2Fdigital_logic%3Akmap-standard-sop-123810) — Digital Logic, chapter 3, asked 2×
- [K-map Π(0,4,5,8,9,11,15), standard SOP](https://www.bctengineer.com/question.md?key=semester_3%2Fdigital_logic%3Akmap-maxterm-sop-0458) — Digital Logic, chapter 3, asked 2×
- [K-map Σ(0,1,4,8,10,11,12)+d, standard minterm](https://www.bctengineer.com/question.md?key=semester_3%2Fdigital_logic%3Akmap-standard-minterm-01248) — Digital Logic, chapter 3
- [MapReduce worker failure tolerance](https://www.bctengineer.com/question.md?key=semester_8%2Fbig_data_technologies%3Amapreduce-fault-tolerance) — Big Data Technologies, chapter 3