Types of NoSQL Databases: Everything You Need to Know About Them
NoSQL or Not Only SQL is a renowned database management system (DBMS) that manages a large volume of unstructured or semi-structured data. Since it eliminates various limitations of conventional relational databases, the NoSQL database has become popular. Google, Facebook, Amazon, and Netflix are some reputable companies that use NoSQL. This blog makes you aware of different types of NoSQL databases. In addition, you will learn their features. Before we move further, let’s find out how NoSQL is diverse from SQL. SQL vs. NoSQL Databases: Quick Comparison Type SQL databases are Relational Databases, while NoSQL databases are known as non-relational databases. Language of Query SQL databases use a Structured Query Language to do jobs like Delete, Select, Update, and Insert. On the other hand, NoSQL has its query language for manipulating data. NoSQL works on a framework or API, depending on the type of database. Expandability Traditional SQL databases are vertically scalable. You can enhance their performance by upgrading hardware. On the contrary, NoSQL databases are horizontally scalable from the ground up. Consequently, they are better at handling large amounts of data and traffic. Property Followed SQL follows ACID (Atomicity, Consistency, Isolation, and Durability) transactions when it comes to managing data integrity. NoSQL databases use the CAP theorem (Consistency, Availability, and Partition Tolerance). Types of NoSQL Databases We can categorize NoSQL databases into the following 4 types. Each has its pros and limitations. You can choose them based on your requirements. Let us learn about them in detail. Key Value Pair Database Key-Value Pair Database is one of the simplest types of NoSQL Databases. It is a non-relational database storing data elements in key-value pairs. Key-Value Pair Database can handle heavy loads of data. It stores data as a hash map and has two columns, i.e., the Key and the Value. Each database key is different, while the value can be String, Binary Large Objects, or JavaScript Object Notation. The three major features of the Key Value Pair Database are speed, straightforwardness, and scalability. Generally, this type of database is used for creating dictionaries, user profiles, user preferences, etc. Graph-Based Database The graph-based database helps users store entities and relations between those entities. Commonly, this database is used to store data on social networking websites, fraud detection systems, healthcare networks, and more. The graph-based database stores the data as a node. The connections between nodes are known as edges. Every edge and node has a different identifier. The database allows users to find the relationship between the data with the help of links. Unlike relational databases, graph-based databases are multi-relational. A few well-known graph-based databases are Flock DB, Neo4J, Infinite Graph, etc. All-in-all, we can say that a graph-based database stores, manages, and queries data as a graph structure. Column Oriented Database Column Oriented Database is a non-relational database. The database lets you store data in rows and read it row by row. It is like a collection of columns like we see in a table. Each column stores one type of information. The database reads and retrieves the data at high speed. You can run analytics on a limited number of columns to read those columns without consuming memory on unwanted data. Column Oriented Database performs queries like Count, SUM, AVG, and MIN quite quickly. Therefore, the database is used for analytics and reporting, data warehousing, and library card catalogs. Document-Oriented Database A document-oriented database is one of the prominent types of NoSQL databases. It stores and manages data like we organize documents in the real world. Although the data is stored and retrieved as a key-value pair, the value is stored as a document. The database uses the JSON, XML, or BSON documents to store the data. Users can store and retrieve documents from their networks in a form that is closer to the data objects. Therefore, negligible translation is needed to access and use data in an application. Document-Oriented Database supports flexible schema, scalability, and quick retrieval. MongoDB and Couchbase are two fine examples of these databases. This database is used in CMS (Content Management Systems), E-commerce websites, gaming applications, collaboration tools, etc. So these are four types of NoSQL databases. Let’s find out why this database system is getting popular. Features of NoSQL NoSQL has several advancements over traditional databases. We have listed a few significant ones. Compatible with Multiple Data Models Like relational databases, NoSQL is not strict. It can handle multiple data models. Additionally, the database can manage structured, semi-structured, and unstructured data with the same speed. Schema Flexibility Unlike conventional database systems, Not SQL databases do not require a fixed schema. It supports relaxed schemas. NoSQL is capable of managing different data formats and structures. As it does not have a strict predefined schema, it permits changes in data models. Scalable As mentioned above, the NoSQL database is scalable. Users can scale it horizontally by adding more modes and servers. Consequently, it is suitable for websites and web applications with continuously growing data. Excellent Uptime NoSQL databases have excellent uptime. They support serverless architecture and create multiple copies of data on various nodes. Consequently, businesses manage their database smoothly with minimal downtime. If one note breaks down, another takes its place and gives access to the data copy. Examples of NoSQL Now you know the different types of NoSQL databases and their uses. Below are some examples of them. Document Database MongoDB is a well-known document-oriented database. It stores data in JSON-like documents. MongoDB is popular for its scalability and flexibility. Column Database Apache Cassandra is a well-known column-based database system that handles large amounts of data across different commodity servers. Graph Database Amazon Neptune is a managed graph database service by AWS. It can work with both RDF graph and property graph models. Key-Value Database Amazon DynamoDB is a database service that provides high uptime and low-latency key-value storage. This service from Amazon Web Service is the epitome of a Key-Value type database. Conclusion Various types of NoSQL databases are a crucial
Kubernetes Persistent Volumes — 5 Detailed Steps to Create PVs
If you want to persist data in Kubernetes, you may utilize the readable and writable disk space available in Pods as a convenient option. But one thing you must know is that the disk space depends on the lifecycle of Pod. Unsurprisingly, your application development process features independent storage available for every node and can handle cluster crashes. Kubernetes Persistent Volumes got your back with their independent lifecycle and great compatibility for stateful applications. This article will lead you to 5 extensive steps to create and implement persistent volumes in your cluster. Before that, let’s dig down to know what exactly persistent volumes in Kubernetes are along with some important terms! Persistent Volumes in Kubernetes A Kubernetes Persistent Volume is a provisioned storage in a cluster and works as a cluster resource. It’s a volume plugin for Kubernetes with an independent lifecycle and no dependency on the existence of a particular pod. Unlike containers, you can read, write and manage your databases without worrying about disk crashes because of restart or termination of the pod. As a shared unit, all the containers in a pod can access the PV and can restore the database even if an individual container crashes. Here are some important terms you must know! Access Modes The accessModes represent the nodes and pods that can access the volume. The field ReadWriteOnce defines every pod having access to read and write the data in a single mode. If you’re using Kubernetes v1.22, you can read or write access on a single node using ReadWriteOncePod. Volume Mode The volumeMode field is mounting functionality of volume into the pods based on a pre-set directory. It defines the behaviour of volume in each Filesystem of a pod. Alternatively, you can use a volume as a raw block storage without any configuration with a Block field. Storage Classes As the name describes, storage classes are the different storage types you can use according to the hosting environment of your cluster. For instance, you can choose azurefile-csi for Microsoft Azure Kubernetes (AKS) clusters while do-block-storage is great for DigitalOcean Managed Kubernetes. Creating a Persistent Volume Step 1: YAML file The process of creating Kubernetes persistent volumes starts with creating a YAML file. The storage configuration represents a simple persistent volume of 1 Gi capacity. Here’s how you can create a YAML file for your PV in Kubernetes: apiVersion: v1 kind: PersistentVolume metadata: name: example-pv spec: accessModes: ReadWriteOnce capacity: storage: 1Gi storageClassName: standard volumeMode: Filesystem Step 2: Adding Volume to the Cluster Once you have created the Persistent Volume, you can add your new persistent volume to your cluster. We recommend using Kubectl for this to make it easier. To add new persistent volume, run: $ kubectl apply -f pv.yaml If you see the following error message while running the command, The PersistentVolume "example-pv" is invalid: spec: Required value: must specify a volume type Try using dynamic volume creation which will automatically create a persistent volume whenever it’s used. That’s because the cloud providers usually restrict allocating inactive storage in the cluster and dynamic volume can be your good-to-go option. Step 3: Linking Volumes to Pods Linking PVs with the pods requires the request to read/write files in a volume. Here the Persistent Volume Claim (PVC) can get you access to the example-pv volume. Let’s see how an example volume claim looks like! apiVersion: v1 kind: PersistentVolumeClaim metadata: name: example-pvc spec: storageClassName: "" volumeName: example-pv As discussed above, you may need dynamic volume creation in some scenarios. You can request a claim for that in the way mentioned below. apiVersion: v1 kind: PersistentVolumeClaim metadata: name: example-pvc spec: accessModes: – ReadWriteOnce resources: requests: storage: 1Gi storageClassName: standard Now, you have unlocked accessModes and storageClassName fields after the claim. All you need to do is to apply the claim to your cluster using Kubectl. Run the following command to quickly apply the claim to your cluster. $ kubectl apply -f pvc.yaml persistentvolumeclaim/example-pvc created In the last, use the volumes and volumeMount fields to link the claim to your pods. This will add pv to your containers section of the manifest and make the files overlive the container instances. To link the claim, run: apiVersion: v1 kind: Pod metadata: name: pod-with-pvc spec: containers: name: pvc-container image: nginx:latest volumeMounts: – mountPath: /pv-mount name: pv volumes: – name: pv persistentVolumeClaim: claimName: example-pvc Step 4: Demonstrating Persistence In the demonstration, you can verify the behaviour of PV in different scenarios. Let’s take a quick example for better understanding. Get a shell to the pod: $ kubectl exec –stdin –tty pod-with-pvc — sh Write a file to the /pv-mount directory mounted to: $ echo "This file is persisted" > /pv-mount/demo Detach the file from the container: $ exit Delete the pod using kubectl: $ kubectl delete pods/pod-with-pvc pod "pod-with-pvc" deleted Recreate the pod: $ kubectl apply -f pvc-pod.yaml pod/pod-with-pvc created Get a shell to the container and read the file: $ kubectl exec –stdin –tty pod-with-pvc — sh $ cat /pv-mount/demo This file is persisted Step 5: Managing Persistent Volumes Kubectl allows you to manage your Kubernetes Persistent Volumes whether you want to retrieve a list or remove a volume. To retrieve a list of PVs, run: $ kubectl get pv NAME CAPACITY ACCESS MODES RECLAIM POLICY STATUS CLAIM STORAGECLASS REASON AGE pvc-f90a46bd-fac0-4cb5-b020-18b3e74dd3b6 1Gi RWO Delete Bound pv-demo/example-pvc do-block-storage 7m52s Review persistent volume claims: $ kubectl get pvc NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE example-pvc Bound pvc-f90a46bd-fac0-4cb5-b020-18b3e74dd3b6 1Gi RWO do-block-storage 9m Sometimes, a volume or PV claim may show a Pending status as the storage class is yet to provision storage. But you can check what’s slowing down the claim process in object’s event history with describe command. $ kubectl describe pvc example-pvc … Events: Type Reason Age From Message —- —— —- —- ——- Normal Provisioning 9m30s dobs.csi.digitalocean.com_master_68ea6d30-36fe-4f9f-9161-0db299cb0a9c External provisioner is provisioning volume for claim "pv-demo/example-pvc" Normal ProvisioningSucceeded 9m24s dobs.csi.digitalocean.com_master_68ea6d30-36fe-4f9f-9161-0db299cb0a9c Successfully provisioned volume pvc-f90a46bd-fac0-4cb5-b020-18b3e74dd3b6 Conclusion: By combining Kubernetes and Persistent Volumes, you can effectively and easily
