HBase Pros & Cons

Disadvantages of HBase

After we have discussed the various advantages of HBase, let us analyse some of the shortcomings of HBase. 

Replacement is not easy.

There are many features of the traditional model which are not supported by the HBase. Because of this, it is not considered to be the best replacement for traditional models.

Not like SQL 

Since HBase does not support SQL structure, hence we cannot use it like SQL. Querying is complex and not as simple as that in SQL. HBase does not contain any query optimizer.

No transaction feature

The HBase does not provide a mechanism to start a transaction and roll back. Therefore it is always suggested to use HBase where we do not require the transaction feature.

It does not supports JOIN operation 

The JOINS are handled in the MapReduce layer. Both normalization and joining are very difficult.

Expensive

It is expensive in terms of hardware requirements and memory blocks allocations. It supports only one default sort per table. It results in unpredictable latencies when we integrate HBase with MapReduce jobs.

No default indexing

Unlike the RDBMS, HBase does not provide the facility of default indexing. One has to manually perform the indexing. Another disadvantage is that there cannot be more than one indexing in the table. Only the row key column acts as a primary key in HBase.

FAQs

1. What are the advantages of HBase?
   It can handle large data sets. It is scalable. It is schema-free. It provides Java API for clients, etc.
    
2. What are the disadvantages of HBase?
   It is not an easy replacement, does not support various RDBMS features, querying is complex, and does not provide default indexing.
    
3. When should we not use HBase?
   It is not optimized for transactional and join operations. So whenever these two operations are to be used frequently, HBase must be avoided.
    
4. Why is querying not optimized in HBase?
   It is not a relational database. It does not support query language because of which querying is not very efficient.
    
5. Where can we use HBase?
   HBase is used in cases where we need random read and write operations and it can perform a number of operations per second on a large data sets. HBase gives strong data consistency.
    

Take this awesome course from coding ninjas.

Key takeaways 

• There are various features of HBase that make it the top choice of users. However, there are some shortcomings.
• HBase can be used to handle large amounts of data in the range of petabytes. It stores large data sets on the top of HDFS file storage and can be used for analytics.
• The searching and processing are much faster in HBase.
• It provides greater flexibility, reliability, and scalability. It is linearly scalable.
• There is no concept of a fixed schema like the RDBMS.
• There are many features of the traditional model which are not supported by the HBase.
• Querying is complex and not as simple as that in SQL. HBase does not contain any query optimizer.
• Unlike the RDBMS, HBase does not provide the facility of default indexing.

Never stop learning. Explore more here.

Happy learning!