Cloud Bigtable; An Introduction

 

Google Cloud Bigtable is a fully managed, highly scalable NoSQL database service for analytical and operational workloads. It provides high performance and predictable low latency, 99.999% availability, and scales to thousands of nodes while leveraging hardware failure handling built into the system. In addition, Bigtable has a simple, logical data model that's easy to use and understand by your application developers.

Google Cloud Bigtable is a petabyte-scale, fully managed NoSQL database service for large analytical and operational workloads.

Both Google Cloud Bigtable and DynamoDB are NoSQL, or non-relational, databases. This concept means their data is stored in some format other than 2D tables. Some commonly-used formats for NoSQL databases in general are documents, key-value pairs, graphs, and columns, and with different strengths and tradeoffs.

Google Cloud Bigtable and AWS DynamoDB are both highly-available, scalable, globally distributed and fully-managed serverless NoSQL databases. Both can function as a key-value store, however, DynamoDB additionally supports a document model and Bigtable additionally supports a wide-column a wide-column store. Both offer two consistency levels: eventual consistency, and immediate consistency.

Benefits

1. Fast and performant: use Cloud Bigtable as the storage engine that grows with you from your first gigabyte to petabyte-scale for low-latency applications as well as high throughput data processing and analytics.

2. Seamless scaling and replication: start with a single node per cluster, and scale to hundreds of nodes dynamically supporting peak demand at low latency. Replication also adds high availability and workload isolation for live serving apps.

3. Open and integrated: easily connect to the open source ecosystem with the Apache HBase API. Build data-driven applications faster with seamless integrations with Hadoop, Dataflow, Dataproc, and BigQuery.

Features of Google Cloud Bigtable

1. High throughput at low latency: Bigtable is best for storing extremely large volumes of data in a key-value store and provides high read and write throughput with minimal latency, allowing for quick access to massive amounts of data. Throughput increases linearly, therefore adding more Bigtable nodes will increase QPS (queries per second). However, this is based on the same infrastructure that runs Google's billion-user products like Search and Maps.

2. Cluster resizing without downtime: Bigtable throughput can dynamically modify by adding or deleting cluster nodes without restarting, which means you can scale up a Bigtable cluster for a few hours to manage a huge load, then scale back down without any downtime.

3. Automated replication for optimizing any workload: write data once and it automatically replicates where it is necessary, ensuring high availability and separation of read and write workloads. And for assuring consistency, correct data, and synchronizing writes and deletes, no human procedures are necessary. For instances with multicluster routing over three or more regions, a high availability SLA of 99.999% is available (99.9% for single cluster instances).

4. Simple administration: Bigtable performs updates and restarts invisibly, and it maintains excellent data durability automatically. Simply add a second cluster to your instance to duplicate your data, and replication will begin immediately. Moreover, there is no need to worry about replication or regions, simply design your table schemas and Bigtable will take care of the rest.

5. Cluster resizing without downtime: to accommodate a heavy demand, you may raise the size of a Bigtable cluster for a few hours, then lower it again - all without any downtime. Bigtable balances performance across all nodes in your cluster in only a few minutes under load when you modify the size of your cluster.

6. Security: Your Google Cloud project and the Identity and Access Management (IAM) roles that you provide to users manage access to your Bigtable tables. You may use IAM roles to block users from reading from tables, writing to tables, or creating new instances, for example, However, anyone who doesn't have access to your project or an IAM role with valid Bigtable permissions cannot access the tables.

7 Data durability: when you use Bigtable, your data get store on Colossus. This is Google's own extremely durable file system, on storage devices in Google's data centers. However, for using Bigtable, you don't require an HDFS cluster or any other file system. Bigtable keep one copy of your data in Colossus for each cluster in your instance if your instance utilizes replication. Each duplicate is in a distinct zone or area, which improves durability even further.