AWS (Amazon Web Services), Azure and GCP (Google Cloud Platform) are 3 main big players in the cloud market today, however this post is dedicated to a comparison between 2 specific storage solutions provided by AWS & GCP.
AWS, The great cloud eco-system provides Amazon Glacier which serves as a cold storage service for the enterprises. Glacier is highly durable and secure in terms of storage. It also provides integration with other services like Amazon SNS (Simple Notification Service) for notifications. The solution provides storage at minimum retrieval response time of 3 to 5 hours and maximum depending upon the selected retrieval policy.
A user can interact with the glacier as per the given flow. A user uploads the data, gets an identifier for the uploaded data and can request Glacier for retrieval, however the request for retrieval again returns a job id which is further used to download the actual data.
Now the catch here is Amazon Glacier does not provide a way to see the inventory of the data that is there inside a Glacier Vault (an bucket like component inside which files are stored ) on the Management Console as well as it does not provide a way to upload the data from AWS Management Console. In short almost all direct interactions (like upload and retrieval) with AWS Glacier needs programming solutions.
Glacier integration with S3
There is one other way with which you can perform your operations with Glacier, is through Amazon S3. Amazon S3 lets you set the lifecycle rules for the data uploaded into an S3 bucket to be automatically transferred into a Glacier vault.
Google’s Nearline is not an individual service but a storage category which works exactly same as AWS Glacier. However Nearline promises retrieval of the data with an average response time of 3 to 5 seconds.
Moreover Nearline gives advantage by giving users the transparency of their data even inside the cold storage by providing them inventory. It also provides an option to download and upload the data from the web interface itself. The users who are familiar with AWS S3 will have almost no struggle to use ‘Google Storage’ and indeed Nearline.
Looking at this Nearline seems to have an upper hand over Glacier but there are some facts which needs consideration as well.
Nearline has a limit of 4MB/sec for every 1 TB stored over retrieval rate where as Glacier provides a higher rate of retrieval. To conquer this problem Nearline provides an option of On Demand I/O which allows you to increase your retrieval rate, however the option of On Demand I/O is off by default.
So its really a comparison between ‘a sweet and costly apple provided by a big and quality conscious shop’ and ‘a sweeter and cheaper apple provided by a relatively small shop’.