Google Cloud Storage


Introduction to GCS


In this post, we will look at the GCP Storage options and service providings for unstructured data objects. GCS(Google Cloud Storage) forms a bucket to store the objects and provides the flexibility to compute instances. Also the GCS Buckets provide the possibility to share it across multiple instances and zones. We can also use the storage bucket as a medium for transferring and backing up the on-premise instances. 

When you create a bucket to store objects, you can select four types of classes. The performance of the buckets depends on the location and instance which is accessed for and the class type. The types are classified according to storage durations, high availability and pricing. 


Multi-regional storage


Multi-regional storage a.k.a. Geo-redundant storage ensures that to have at least two copies of your data store in geographically separated locations. 
This type is suitable for frequently accessed data. However, the geo-redundancy happens asynchronously across the locations but cloud stores ensure that it copies at least one location as soon as the data uploads. 
This offers 99.95% SLA availability.

Regional storage


This type is suitable for storing the data in the same region and does not offer the redundancy features, as this stores the data in the same region provides better performance throughput. 
It provides 99.90% SLA availability.

Nearline storage


The nearline storage class is appropriate for the data that is accessed rarely, like fetching once in a month. This would be a more suitable option for backups and archives operations, as this offers 99.0% SLA availability. 
This will also incur data retrieval costs for fetching the data which is more than 30days duration. 

Coldline storage


This type is for completely infrequently accessed data like archiving the yearly wise data which has a very rare situation to see it. That is why this offers less storage cost compared to other class types. 
It offers 99.0% SLA availability and incurs retrieval cost for fetching data which is more than 90days duration. 

The tricky part here is, if no storage class selected, the standard storage type will be applied. Standard storage type means multi-regional class if the bucket is in the region where multi-regional supported or regional class if the bucket in the region where only regional class supported. I hope this is clear. 

Create the storage bucket


Let us move on to create the storage bucket from the console. 

gcp services

You might get this warning if the billing option is not enabled. 

gcp storage bucket


Provide a suitable bucket name and also have the estimated billing for storing the objects. For example, if you store 100GB of data and retrieve 10GB, then the monthly cost is around 2.60$ only. 

gcp storage bucket

Choose the store location type. In this case, I just go with a single region,

gcp create bucket


Select the storage class based on the requirement, I have chosen standard for my case,

gcp storage class

gcp storage class


And go on to create the bucket with default settings. That is all I have for this demo. Please do visit other posts in the cloud tab. Thank you. 


Recent Posts