NoSQL: What is Bigtable on GCP?
If scalability and reliability are important to you, BigTable could be your choice, but many NoSQL databases like MongoDB or Cassandra are like this, but what is more important in Google Cloud BigTable?
1. Low latency
According to Google, BigTable offers a consistent sub-10ms latency and this is perfect!
If this is the first time you hear BigTable, do not be surprised, we can understand it very easily, we just need to know a little about Wide Column based Types of NoSql Databases. Let’s talk about it.
Imagine you are a restaurant owner. A strange thing happened today. One of the customers came to the restaurant today and claimed that he had already booked a table for 12 o’clock and you do not remember it. You use the telephone method to handle reservations and save time and other information in your notebook. What you wrote for today is as follows!
Well! How to prepare your report for the customer in the shortest time?! To reduce reporting time, do you just provide a list of phone numbers and a list of booking time, or do you also show the foods and table numbers to the customer? Of course, to convince that customer, you only need the phone number column and reservation time, so you ignore the other columns. this is how wide-column-based works and one of the differences with row-based DB is the same reporting method that MySQL, Postgres, etc. use. this example will help you to know how to design your schema tables.
The result of this feature in BigTable is high-speed, efficient, and reliable data retrieval!
Another thing you need to know is that BigTable can be used to model key-value datasets like google cloud memory store, But for example in the implementation of the leaderboard of games where cache-to-DB consistency starts to matter. Otherwise, memory store is better!
2. Replication & Durability
Replication for Cloud Bigtable enables you to increase the availability and durability of your data by copying it across multiple regions and zones within the same region. Its high availability for live serving apps and workload isolation and also Replicated clusters in different regions typically have higher replication latency than replicated clusters in the same region.
In MongoDBAtlas also can enable cross-region replication for multi-region fault tolerance and fast, responsive reads.
3. Massive Storage Engine
Cloud Bigtable stores data in massively scalable tables, If you are looking to store information in large volumes like Stream processing/analytics, and machine-learning it’s one of the best choices. Cloud Bigtable treats all data as raw byte strings for most purposes. Many options like Empty cells in a table do not take up any space or compaction data Increases your productivity in large data!
4. Cluster resizing without downtime
To handle a large load you might increase the size of a Cloud Bigtable cluster. According to Google After you change a cluster’s size, it typically takes just a few minutes under load for Cloud Bigtable to balance performance across all of the nodes in your cluster. So in live serving apps, you can handle a large load without any downtime!
5. Compatible with Google services
As a Google product, BigTable integrates well with other Google Cloud services and platforms
When Should You Use It?
Cloud Bigtable is a sparsely populated table that can scale to billions of rows and thousands of columns. According to Google, BigQuery is great for IoT, and advertising technology. The most important reason you can use BigTable is to deal with a lot of data and process/analyze in the fastest time, such as machine learning and such projects … Like any other column-based database, fast search and high-speed Aggregate responsiveness is one of the advantages of this powerful NoSQL database.
