Participated in Google Cloud OnBoard

My name is Ito, and I work as an infrastructure engineer and public relations officer

I participated in the "Google Cloud OnBoard" held on December 13, 2016

Google Cloud OnBoard

Google Cloud OnBoard was an introductory training session for Google's cloud platform, "Google Cloud Platform" (GCP), and it
was the first time it was held in Japan. It
provided an introduction to the many features of GCP, including use cases and demonstrations,
and allowed participants to hear "what goes on inside Google" from Google employees.

I would like to give you a brief introduction to the contents of Google Cloud OnBoard

Google's way of thinking

Google's philosophy is "services for everyone around the world," and to ensure
that anyone, anywhere in the world can use the same services,
they are required to operate as a single "DataCenter as a Computer," keeping all of Google's data centers worldwide in the same state.

Google's technology is publicly available in some places, so it can be a useful reference (though it's in English):
Research at Google

At the heart of Google's various services is a container management system called Borg, and many services, including BigQuery, run on Borg.
Incidentally, there is an open-source version of this called Kubernetes.

Kubernetes - Production-Grade Container Orchestration

BigQuery has the capability to perform regular expression replacements on 10 billion rows in just under 10 seconds.
Internally, BigQuery works by
splitting the data, storing each section on an HDD, retrieving the data when a query is run, and creating separate containers for each section.
Since disk I/O becomes a bottleneck when running queries, splitting the data into numerous containers enables high-speed analysis.

By the way, I heard that BigQuery doesn't create indexes and instead performs full scans. Apparently,
it's more difficult to create indexes because the data is so large.

developed and used in-house—that is, a "service used all over the world"—really
that they offer GCP as a service
They want people to use the services they offer as PaaS as much as possible, rather than using GCE (or EC2 in the case of AWS), and
to rely on Google's ultra-fast infrastructure so they can focus on creating content.

Low price

As a company that handles various cloud services, price is always a major concern.
GCP uses a "per-minute billing" system instead of the more common "per-hour billing" model.

For all machine types, a minimum usage fee of 10 minutes is charged. For example, even if you run a virtual machine for 2 minutes, you will be charged for 10 minutes of usage.
If the usage exceeds 10 minutes, the instance will be billed in one-minute increments. If the instance runs for 11.25 minutes, you will be charged for 12 minutes of usage.

Additionally, you'll receive a discount just for using the service.
If you use the machine for more than 25% of the month, you'll be charged 80% of the base rate.

Compute Engine automatically provides discounts on every minute of your instance usage if you use it for more than 25% of the month. The more hours you use, the higher the discount, up to a net discount of 30% on instances that run for the entire month

The third pricing feature is "estimated instance."
I think the easiest way to understand this is to look at an image, so I'll borrow an image from the official pricing page.

Within each specific project, if there are instances used for various periods, such as "1 day" or "10 days," the instance usage is
"averaged" to calculate the cost in a way that maximizes the discount.
, if there are two instances used for "15 days," they will be treated as one instance used for a full month, and the automatic renewal discount will be applied.

Reference:Google Compute Engine Pricing | Compute Engine Documentation | Google Cloud Platform

I mean, there are a lot of people

The event was held at the Belle Salle Shibuya venue, and apparently it exceeded initial expectations, with so many people that some had to stand. There
were around 1000 people...!

The lines for the toilets were incredibly long, and so were the lines for coffee and other drinks.
Even though lunch boxes were being distributed at the free event, they were so crowded that it was difficult to get one!

There were all sorts of gifts, and everyone got a T-shirt, so
I thought it was a really luxurious study session.

This Google Cloud OnBoard event was more of an introduction to GCP, so
I'd like to participate in a hands-on seminar that delves deeper into the topic.