Q6: What is the difference between MongoDB and a relational database?
Q7: What are the limitations of MongoDB? How do you handle them?
Q8: What is the difference between Sharding and Partitioning in MongoDB?
Q9: What is the difference between vertical Partitioning and Normalization?
==============================================================================
Q6: What is the difference between MongoDB and a relational database?
Q6: What is the difference between MongoDB and a relational database?
Answer:
MongoDB: schema-less, document-based, horizontal scale-out, flexible data model.
Relational DB: strict schema, table-based, vertical scaling, strong consistency.
Architecturally, MongoDB makes sense where:
Data structures evolve frequently.
Reads and writes are distributed at scale.
Complex joins aren’t dominant.
However, for transactional consistency-heavy use cases (banking, financial core systems), I’d still consider RDBMS or a hybrid approach (polyglot persistence).
==============================================================================
Q7: What are the limitations of MongoDB? How do you handle them?
Answer:
Limitations:
- Less suited for heavy relational joins.
- Document size limit (16 MB).
- Write amplification with large indexes.
- Transactions come with performance overhead.
As an architect, I’d mitigate by:
- Choosing polyglot persistence (e.g., use MongoDB for product catalog, PostgreSQL for payments).
- Offloading complex aggregations to data warehouses.
- Designing sharding and replication for scale and resilience.
==============================================================================
Q8: What is the difference between Sharding and Partitioning in MongoDB?
Answer:
| Aspect | Partitioning | Sharding |
|---|---|---|
| Definition | Logical division of data within a single node or database, often to separate workloads or improve query performance. | Horizontal scaling across multiple servers (nodes), where each shard holds a subset of the data. |
| Scope | Usually internal to a single instance/database. | Distributed across multiple nodes in a cluster. |
| Purpose | Improve manageability, optimize queries, or segregate data. | Scale writes and reads horizontally, handle very large datasets, ensure high availability. |
| Implementation in MongoDB | MongoDB doesn’t have explicit “partitioning” in the classic RDBMS sense. You can simulate partitioning using collections or databases per category. | Implemented via sharding, which uses a shard key to distribute data across shards. Includes config servers and mongos routers to manage routing. |
| Fault tolerance | Limited — if node fails, data is lost unless replicated. | High — usually combined with replica sets per shard for redundancy. |
| Complexity | Low — simple logical separation. | High — requires careful shard key selection, balancing, and operational management. |
==============================================================================
Q9: What is the difference between vertical Partitioning, Horizontal Partioning and Normalization?
Answer:
They look similar at first glance, but they’re not the same thing. Let’s untangle them.
Vertical Partitioning
- Splitting a table into multiple tables based on columns.
- Done mostly for performance, manageability, or storage optimization.
- Example:
Normalization
- A database design process to reduce data redundancy and improve data integrity.
- It breaks down a table into multiple tables based on functional dependencies.
- Example: A Customer table storing city names directly leads to repeating the same city thousands of times.
- Normalization moves city info into a City table and links via foreign key.
The Key Difference
- Vertical Partitioning is a physical optimization technique (to improve performance and manageability).
- Normalization is a logical design principle (to ensure consistency and reduce redundancy).
- Sometimes, vertical partitioning looks like normalization because both result in more tables. But their motivation is different:
- If you’re doing it to enforce data integrity, that’s normalization.
- If you’re doing it to speed up queries or reduce I/O, that’s vertical partitioning.
==============================================================================
==============================================================================
