Google Cloud
Google Cloud

Ways to Use Google’s Free Tier

Use App Engine:

 

Google App Engine remains one of the best ways to spin up a web application without having to worry about all the details of how to deploy or scale it. Nearly everything is automated, so new instances will be launched if the load increases. App Engine comes with 28 “instance hours” per day – meaning your core application will run free for 24 hours a day and can even scale up for four additional hours if there is a spike in demand.

 

Consolidate Service Calls:

 

There is some room to add extras if you are careful. The limits on serverless calls are based on the number of individual requests, not on complexity. You can pack more actions and more results into each exchange by bundling all data operations into one larger batch. So you can deliver frivolous tricks like stock prices, but only if you squeeze those extra few bytes into the very core packets. Just keep in mind that Google counts memory used and computation time. Your functions cannot exceed 400,000 GB-seconds of memory and 200,000 GHz of computing time.

 

Use Local Storage:

 

Modern web APIs provide a number of good places to store information. There is the perfectly good old-fashioned cookie, which is limited to four kilobytes. The Web Storage API is a document-based key-value system that stores at least five megabytes of data, and some browsers will retain up to 10 megabytes. IndexedDB offers a richer set of features such as database cursors and indexes that will speed through data that is stored frequently, largely without limits.

 

The more data you store locally on the user’s device, the less you need to use precious server-side storage. This can also mean faster responses and far less bandwidth dedicated to transferring countless copies of data to your server. There will, however, be issues when users switch devices, as the data may not be synchronized. Just make sure the important details remain consistent.

 

Look for Hidden Deals:

 

Google maintains a helpful page that summarizes all “always free” products, but if you browse around you will find plenty of free services that do not appear on that list. Google Maps, for example, offers “$200 worth of free monthly usage.” Google Docs and a handful of other APIs are always free.

 

Use G Suite:

 

Many G Suite products — including Docs, Sheets, and Drive — are billed separately, and users can either get them free through their GMail account or have their business pay for them as a bundle. Instead of building an application with built-in reports, simply write the data to a spreadsheet and share it. Spreadsheets are powerful enough to embed charts and plots just like any dashboard. If you build a web app, you will need to replicate the account quotas and your data to handle interactive requests. But if you simply create a document in Google Docs for your report, you offload most of the work onto Google’s infrastructure.

 

Drop the Gimmicks:

 

Some features of modern web applications are rather unnecessary. Does your banking app really need stock prices? Do you need to embed the local time or temperature? Do you need to embed the latest tweets or Instagram photos? No. Drop all these add-ons, because each one means another connection to your server infrastructure and that eats into your free allowance. The product design team may dream big, but you can tell them, “No!”

 

Be Careful with New Options:

 

Some of the cooler tools for adding AI services to your stack offer decent limits for experimentation. The AutoML Video service will let you train your own machine-learning model on video feeds for 40 hours each month before charges kick in. The Scheduled Data service will grind through your rows and rows of information on a free node for six hours. This gives you enough rope to try out or build basic models, but beware. It would be risky to automate the process so that every user can trigger a large machine-learning job.

Keep Costs in Perspective:

 

It is easy to take this game to the extreme and turn your application architecture into a Rube Goldberg machine just to save a little more money. It is important to remember that stepping up from the free tier to a paying customer is often a very small step on Google Cloud. While there are many free services on the internet that jump from free to thousands of dollars with a single click, Google’s services are generally not priced that way.

 

After two million free Cloud Function invocations, the next call costs a whopping $0.0000004. That is just 40 cents per million. If you dig around in the sock drawer, you should be able to cover a few extra million with very little trouble.

The pricing table is generous enough that you will not have a heart attack when you leave the free zone. If your application needs a few extra million of this or that, you can probably cover it. The important lesson is that keeping computational load low will translate into smaller bills and faster responses.

 

Read More Also About:

How to Get the Most Out of Google Cloud’s Free Tier

Meta Faces the Collapse of Its Cryptocurrency “Meta”

Used Face Masks: From Harmful Waste to Useful Batteries