Greg Brown
Greg Brown
Apr 18, 2019 9 min read

Google Cloud Platform and Applications

thumbnail for this post

Google is battling tooth and nail with alternative high-profile cloud platforms. Their impressive list of benefits continues to gain strength, as more businesses crave online functionality. Google knows a thing or two about handling vast amounts of data. Google developed cloud services on the same infrastructure as their internal end-user products. At present, there are over 90 separate products available on the platform.

Google breaks down its products into seven categories, with each having an extensive set of products. 1) Compute 2) Storage 3) Networking 4) Stackdriver 5) Tools 6) Big Data 7) Artificial Intelligence. The platform is continually evolving.

In a recent survey, SADA found, 84% of IT managers employ cloud services rather than traditional data centers. Some of the largest companies in the world are using Google’s Cloud including Apple, eBay, and TechCrunch. When it is time to choose an infrastructure partner Google’s Cloud Platform has several distinct advantages.

  1. GCP pricing: Google’s Cloud only charges in one-minute increments, which means consumers are just paying for the computing power they need. Startups should use the initial credits and no up-front for long-running workloads.
  2. Google Cloud is one of the largest networks in the world and very fast. A distinct advantage of Google over other players, a privately distributed backbone between data centers the world over. Their 9,000 km trans-pacific cable is the highest capacity undersea cable ever built. Google is the absolute leader in the rapid deployment of fiber-optic and servers. With speed comes, lower costs on more data.
  3. Security is state-of-the art at Google Cloud. The platform has been dealing with web security for over 15 years with their retail products. Among the features: data is encrypted and a commitment to enterprise security.
  4. Constant expansion is the name of the game at Google Cloud. Google Cloud announced they have surpassed the billion-user products mark. With such phenomenal growth, the company opened eight new cloud regions to meet demand.

Three significant services on the Google Cloud Platform are Kubernetes, Cloud Firestore, and Cloud Pub/Sub.

Kubernetes

Kubernetes is an open sourced engine, where developers automate application deployment in a containerized environment. Released in 2014, Kubernetes has a robust ecosystem of add-ons and support systems.

Deployment is by containers based on an operating system rather than hardware. Kubernetes insulate containers from one another and the host. Each container houses unique file systems, resources, and each container holds a specific application.

Kubernetes Deployment

Kubernetes is a contemporary design of deployment, where applications are attached to clusters without connection to the original host. Other deployment models offer less flexibility because applications are tied to a host machine. Kubernetes allows a developer to deploy containerized applications in clusters.

Kubernetes administration is a mix of locations and resources. Developers can use Google Cloud, from the left side menu, to deploy Kubernetes Clusters. Pre-packaged Kubernetes solutions with dozens of categories are available in the Cloud marketplace. Use Google Clouds Compute Engine and deploy large Kubernetes Clusters with unlimited CPUs in either Linux or Windows.

A significant facet of the Kubernetes engine is REST API. Every connection between components is based upon REST API calls. Kubernetes is looked upon as an API object and an entry in the API. Kubernetes is flexible and supports direct communication by a command-line interface. Each of those resources uses the API. If the user wants to develop application clusters, Kubernetes provides an extensive API library.

Kubernetes is not just another development platform. The engine supports the formation of containerized applications at scale. Kubernetes is portable and self-healing, affording a more natural control. Several vital concepts make Kubernetes an exceptional development network. 1) Nodes are the structures of the system, yielding items of information like conditions and capacity. 2) Pods are one or more containers and a place to hold data. 3) Deployments are the means to update a pod and determine the state of a Kubernetes cluster. 4) Services are connections between pods.

Kubernetes is a tremendous open sourced tool for creating applications at scale across a wide variety of hosts and operating systems.

Google’s cloud platform is more than just a hosting platform; it is a launching point of interconnectivity expanding to every intersection of the internet. Scaling is extraordinary for every application. Google’s innovative technologies alter the way enterprises handle and interpret data. Platform tools are quick to deploy with plenty of virtual instances of additional tools for testing.

Cloud Firestore

Database technologies have expanded exponentially to keep up with billions of data points. Companies now use intelligence as a competitive edge. Experts agree to remain relevant; companies need a higher level of data integration with existing business operations. Database technologies concentrate on individual business needs rather than a row and column procedure.

Cloud Firestore is another highly regarded tool available on the Google Cloud Platform. Firestore is a NoSQL (not only SQL) database scheme constructed for specialized development and scaling. There are correlations with a structured query database but with unique variations. Cloud Firestore is real time, assisting in the development of responsive applications. Firestore is a cloud-hosted solution.

Firebase is an application development platform acquired by Google in 2014. Firebase is an excellent companion when developing a database with Firestore. The service produces an API to integrate data across platforms. Several tools are built into Firebase to expand functionality for mobile and web applications.

Firestore is native to the cloud, meaning higher levels of automatic scaling across varied platforms. The Firestore database does not rely on a server, making the program flexible. In addition, users can talk to Firestore from a mobile client.

Data is kept in documents allowing for value mapping, and each record can have several nested rules. Files are then saved to collections and sub-collections for a hierarchal arrangement. Here are a few of the essentials with Cloud Firestore.

Adding Data in Firestore

  • Add Data: There are three techniques to add data to Firestore. It is best to design precisely what your expected output should be. It will save time later on. Firestore allows the user to deposit a broad array of data inside a document including names, Boolean and nested objects. The following are for individual chunks of data, and this guide shows bulk writing. 1) Add specific data points to a document with a unique identifier. 2) Add a new record to an earlier existing collection of documents and let Firestore generate the ID. 3) Add a blank document to a collection, add data later.
  • Updating and Deleting Data: When configuring an update to data within a document, it is best to learn, add a server timestamp to each update. Updates can be applied to an entire document or individual fields within a document. Query Firestore
  • Call and Listen: There are two alternatives to query data from documents, collections, and sub-collections. 1) The call method by default will fetch the freshest information from a record. A user can define whether the data extracted is live or from an off-line cache. 2) With the listener technique, Firestore sends a snapshot every time a document changes.
Security

Firestore has a robust security system built into the database, applying two unique concepts. 1) Firestore uses a rules-based protocol. The rules encompass serverless authentication, approval, and data visualization. 2) Server clients use the Cloud Identity and Access Management protocol.

Functionality

Cloud Functions and REST API: The extraordinary feature set of Cloud Firestore is its ability to operate without server intervention, exclusively in a cloud environment. However, adding Node.js, the user can add server functionality if needed. Developers can enhance the functionality of Cloud Firestore with the addition of native client libraries. Users can call the REST API for specific situations such as a resource-constrained environment or automating the database. Seamless integration between GCP and Cloud Firestore make administering the database efficient. The Google Cloud Console allows users to view, edit and delete data and manage indexes.

Cloud Pub/Sub

Google’s Pub/Sub system is a significant pipeline, more than just delivering messages. It is middleware for enterprise analytics and event ingestion. Pub/sub is part of Google’s Stream Analytics solutions. Stream analytics can absorb millions of data points per second, prepare, and store the knowledge for real-time analysis. Stream analytics offers the enterprise several competencies including fraud detection, personalization, and assimilation of third-party pipelines.

Google Pub/Sub

Stream Analytics on Google Cloud Platform includes the following operations, each contributing an elemental building block to messaging and data analysis:

  • Cloud Pub/Sub
  • Cloud Dataflow
  • Big Query
  • Apache Beam
  • Cloud Machine Learning
  • Cloud Big Table

Pub/Sub shines as a messaging system. Gmail’s framework is based on pub/sub. Gmail delivers 500 million messages per second. Pub/Sub has an exceptional capacity to scale.

A few of the basic core concepts:

  • Messaging: data flowing through pub/sub.
  • Topic: name that describes the feed.
  • Subscription: the company that receives the information.
  • Publisher: an identified entity that issues the message.
  • Subscriber: accepts the message.

Scalability is a central strongpoint to the service. Pub/sub uses a horizontal scaling technique. An expansion in the volume of subscriptions or messages is regulated by raising server instances. There are two parts to pub/sub.

  1. Control Plane uniquely distributes messages providing scalability. When clients connect to the service, routers determine what data center produces the shortest network distance for the message. The router has two goals 1) load balance 2) stability. Google uses a consistent hashing variant developed by its research team.
  2. Data Plane receives the message and dispatches the message to clients. The record of a message is complex within the pub/sub service. Steps to message flow are: 1) Publisher sends a message 2) Message goes into storage 3) Pub/Sub acknowledges receipt and ensures delivery 4) Pub/Sub delivers the message to subscriber 5) Subscriber acknowledges the message 6) After confirmation, Pub/Sub deletes the message

Engineers are situated in several Google Data Centers around the world, ensuring the Pub/Sub service remains viable at peak performance times. Google is continually surveying the service to mitigate issues before they involve the consumer. Site Reliability Engineers use extensive monitoring to keep track of behavior. There are assortments of ways the SREs develop metrics for the Pub/Sub service, such as the time to save a particular message.

Google has constructed an impressive ecosystem of individual applications and interdependent systems all linked to the cloud. The platform is cohesive with so many instances of operating systems and third-party products.

One of Google Cloud’s newest offerings is the Google Cloud Client Libraries. A significant advantage to Google’s Cloud Platform is its extensive list of APIs. Companies are moving entire hosting and web applications to Google’s cloud, because of redundancy, server control, and the ability to tweak every setting. Google’s API directory is an impressive element to their service. Connecting to alternative platforms such as Office 365 in real time revamps the way managers conduct their office. Google is driving a significant push into artificial intelligence and machine learning. The portfolio of operations with Machine Learning is transforming predictive analytics.

comments powered by Disqus