interview questions on cassandra - Wikitechy

What is the difference between Cassandra and Hadoop ?

Editor — Tue, 13 Jul 2021 05:44:49 +0000

Parameters	Cassandra	Hadoop
CAP theorem	AP	CP
Architecture	All nodes are same in Peer-to-Peer Distributed architecture.	Master/Slave. Name Node act as Master and data node act as worker node.
Read and Write Design	Read and write anywhere model.	Write once read many access models.
Area of utilization	Real time online transactional processing	Batch-oriented analytical solutions.
Mode of accessing data	Cassandra query language and Command line interface tools.	Map/Reduce for read/write operations.
Data storage model	Keys space column family to store the data and introduces primary and secondary indexes for high availability of data	File system. Large files are destroyed into small piece and copy on many data node
Fault tolerance	High availability- NO SPOF cluster all nodes are same and experienced to handle the access	when master node is down if the Single point of failure Vulnerable to be failed.
Storage schema	Combines schema from Google big table and Amazon Dynamo	Physical file system schema.
Communication	Gossip protocol	RPC/TCP and UDP

Editor — Tue, 13 Jul 2021 05:03:39 +0000

A primary index is global, although a secondary index is local.
To query a users by their secondary indexed value or their email every machine has to query its self target of users.
One query, five reads from disk.

Using SQL,after defining a table you can create an index on a column.
You can also index a collection column.
Secondary indexes are used to query a table using a column that is not general query table.
Secondary indexes are difficult to use and can impact performance greatly.

The index table is stored on every node in a cluster, so a query involving a secondary index if multiple nodes are accessed it can quickly become a performance nightmare.
A general rule of thumb is to index a column with low cardinality of few data.
Before creating an index, be aware of when and when not to create an index.

Editor — Tue, 13 Jul 2021 04:54:18 +0000

Cassandra database is distrubuted over multiple machines that perform together.
The outermost container is known as the Cluster. For failure handling, every node have a replica, and in case of a failure, the replica takes charge.
Cassandra order the nodes in a cluster, in a ring format, and assigns data to them.
The data model of Cassandra is significantly various from normally what can we see in an Relational Database Management System.

The Group of independent servers (usually in close proximity to one another) interconnected via network to work as one centralized data processing resource.
Clusters are working for multiple complex instructions by distributing workload across all connected servers.
Clustering develops the system’s accessibility to users, its aggregate performance, and overall tolerance to faults and component failures.
A failed server is automatically shut down and its users are switched immediately to the other servers.

The basic attributes of a Key space in Cassandra are,

It represent the cluster contain the number of machines that will collect duplicate of the same data.

It is nothing but the strategy to point replicas in the ring.
We have strategies such as simple strategy (rack-aware strategy), old network topology strategy (rack-aware strategy), and network topology strategy (datacentre-shared strategy).

Key space is a edge for a order of one or more column groups.
A column groups, in turn, is a edge of a order of rows. Each row have a ordered columns.
Column families define the structure of your data.
Each keyspace contain at least one and often many column families.

Editor — Tue, 13 Jul 2021 04:40:35 +0000

Apache Cassandra is a simple and open-source provides NoSQL DBMS build to handle huge volume of data against more fabric servers.It providing high availability with no single point of failure.
The performance of Cassandra place a high value. In 2012, University of Toronto researchers studying NoSQL systems resulted that “Scalability defined there is a clear winner throughout our experiments.Cassandra achieves the largest throughput for the large number of nodes in all experiments” although “it determine at the price of high write and read latencies.”

Cluster contain every node has the same role.
There is no single point of failure.
To provides data across the cluster so every node have various data, but there is no master as every node can service if any request.

Replication strategies are configurable.
Cassandra is built as a distributed system, for deployment of maximum numbers of nodes across many data centers.
Key features of Cassandra’s distributed architecture are specifically change for multiple-data center deployment, for redundancy, for fail over and disaster recovery.

As new machines are added read and write throughput both increase linearly, with the achieve of no halt or interrupt to applications.

Data is automatically replicated to many nodes for fault-tolerance. Replication across multiple data centers is maintained. Failed nodes can be changed with no downtime.

Writes and reads gives a adjustable level of consistency, all the way from “writes never fail” to “block for all replicas to be readable”, with the quorum level in the middle.