In this video, you will learn about using Object Lifecycle policies to control how objects are kept inside a bucket. As you are aware, when you create a bucket, you choose its default here. But if you have a bucket created, we know objects can be there in any tier depending on how you manage the objects. In an Archive tier-based bucket, you can restore objects, which means objects are getting pushed into Standard tier for a certain amount of time. Whereas if you created a bucket of Standard tier, we know we can archive the objects or push objects into Infrequent tier, etc. by changing the Storage tier update option from a given object. But why do we have this option of managing lifecycle of object? Given the construct of the different tiers, we know we can use the three tiers to take care of availability from immediate perspective or restore and access it for objects. There is also a cost difference between the various tiers. Now consider an example wherein I want to have a bucket wherein whenever I upload objects, I want them to go and sit in Standard tier, and I want the objects to be there for 100 days. The assumption is, for the first 100 days since an object is uploaded, maybe it will be frequently accessed and downloaded. But after an object reaches 100 days, I want it to be in Archive tier for the next 200 days, wherein I know I need to keep any object for my reference or archival purposes for another 200 days, but I may not be accessing it regularly. This could be a government's or a regulatory requirement that I need to take care, wherein by keeping the object in Archive tier, I can save some money because it is only one-tenth the cost of Standard tier. Under a situation where I want to access it, we know that we can restore the object and download it. But if this has to be done manually, it is a very effort-intensive task, because every day somebody has to go and look into every bucket, identify which bucket has what type of lifecycle to be managed and go and archive objects and do them. Or you can automate it using Lifecycle policies for your objects. How do we do that? If you go to your bucket, you have the ability to go and create Lifecycle Policy rules, wherein based on the policy you create, you can set an object-level to automatically move it to Archive or Infrequent tier or delete them after a certain number of days. For example, I might add a rule to say archive objects after the 100 days they have been in Standard tier. I might add another rule to say, delete the files after 200 days, so on. As you can see here, you can also include prefixes or filters through which you can specify for which types of objects in terms of the prefix or the naming pattern that is used and create these rules. You have multipart upload, which is a feature which we will talk about in the next video. You can create as many rules. If I'm right, around 1,000 rules per bucket can be available as Lifecycle policy rule. But who is doing this action of moving to Archive or moving to Infrequent or deleting an object? It is done by the service Object Storage. To make these policies work in the tenancy, an IAM policy is required, which will enable the Object Storage service to be able to move. You need to grant the service, the ability to manage objects. Only then these rules will work. That's about Object Lifecycle rules which one can use to manage automatically archiving and deleting files within the bucket so that you disable manual intervention, or rather you don't need to invest in efforts to do that. Whenever you have object Lifecycle Policy rules, there could be multiple rules. When you use the web interface, you can go and edit the rule, create a rule. If you are using CLI and rest APIs and updating the Lifecycle Policy, you need to give the entire bunch of rules. Let's say there are nine rules already available and you want to add two more rules, when you update the policy using APIs or CLI, you need to upload the entire new set of 11 rules, nine old rules plus two new rules and update the policy. You can't do piecemeal or part-by-part edit. You have to upload the entire policy when you want to make these changes. That's about Lifecycle Policy that you can manage for the other objects within a bucket.
