Google Cloud Platform – Google Cloud Storage buckets

How to Setup a Node.js Application Using PM2

For applications that do not have low latency requirements, cloud storage buckets are the answer. Google Cloud Storage buckets form an object storage system that offers a very flexible, reliable, and scalable storage option for a compute engine instance. Cloud storage buckets also make it possible to share instance data across multiple instances and zones. You can also use cloud storage buckets as a common storage between your on-premise instances, as well as other cloud service providers. This logo represents Google Cloud Storage buckets:

Performance of cloud storage buckets depends on the location of the bucket relative to your instance and also the storage class that you select. There are four storage classes that you can select your cloud storage bucket from. The four storage classes are multi-regional storage, regional storage, nearline storage, and coldline storage. While all storage classes offer the same throughput and low latency, they differ primarily in their availability, minimum storage durations, and pricing.

To use a cloud storage bucket in an instance, you need to use the Google Cloud Storage Filesystem in Userspace (FUSE) tool, which is an open source adapter for mounting cloud storage buckets. The mounted bucket behaves as a persistent disk and has a higher latency.

Multi-regional storage is geo-redundant storage that is appropriate for frequently accessed data. The geo-redundant feature ensures that cloud storage stores your data redundantly in at least two geographically separate locations. It is important to remember that geo-redundancy occurs asynchronously but cloud storage ensures that you have at least one copy redundant as soon as you upload your data. This storage class also offers a 99.95% availability SLA. Note that multi-regional storage is available in select locations.

The regional storage class is appropriate for storing data in the same regional location and does not offer the redundancy features that multi-regional storage does. Having data stored within a region results in better performance. Regional storage offers a 99.90% availability SLA.

The nearline storage class is ideal for storing data that is accessed rarely, for example, once a month. Ideally, backups and archival storage are some of the well known use cases for the nearline storage class. The nearline storage class offers a 99.0% availability SLA, and also carries a data retrieval costs along with a 30-day minimum storage duration.

While nearline storage class is ideal for infrequently accessed data, coldline storage is ideal for rarely accessed data such as archives and archival backups. Coldline storage offers long term archival storage at the lowest cost. Coldline storage offers a 99.0% availability SLA and data retrieval costs along with a 90-day minimum storage duration:

If no storage class is selected, the default storage class for the bucket is standard storage. This standard storage class is equal to a multi-regional storage class if the bucket is in a location where multi-regional storage is supported. Standard storage is equivalent to regional storage when the bucket is in a regional location.
  1. To create a bucket, select Storage from the side pane and click on Browser. Select Create to create a bucket:
You can even transfer data from your Amazon S3 instance or from any other HTTP/HTTPS instance or third party storage buckets into a Google Cloud Storage bucket. This option is ideal for data that is less than 20 TB in size.

  1. Select Transfer to configure and initiate transfer of data from third party storage. You can even transfer data from one Google storage bucket to another:

For buckets that are larger than 20 TB in size, a transfer appliance is recommended that can securely migrate data from one bucket to another. Currently, the appliance is provided on a per-request basis, and the form is shown here:

Comments are closed.