Demo: Storage tuning - Amazon Web Services (AWS) Tutorial

- [Instructor] In this demo, we're going to take a look at the storage options available and point out some areas where you might want to consider how you use the actual storage applications or storage services. Starting off with EBS, we have to go under EC2 because EBS volumes only work with EC2 instances. Down on the left, we have volumes, and if I create a volume, I have a number of choices to make really everything's within this screen. Do I want a general purpose drive or do I want to select from one of the IOPS options for more speed? Do I want storage such as colt or magnetic, which is cheaper and perhaps I don't utilize the storage that often and it makes sense. Or do I just want a consistent amount of speed? I'm not interested in IOPS. We have our different choices to consider. We also have to consider where we're placing our volumes that we're creating and you're going to place them in the same availability zone, where your servers are. You could also create a volume from an existing snapshot. So if we actually had a snapshot of a drive you could select that drive and say, well that's what I want to build a drive with the contents from this snapshot. We can also choose to enable encryption using the key management service and merely selecting the default key and moving ahead. So, when we're looking at storage and the EBS volumes we've got quite a few choices and remember EBS E for elastic means I can come back in and make changes later on. As far as the other storage services, all under storage, we have S3 storage and S3 storage allows me to create an S3 bucket. Now, when I create an S3 bucket I have to create a name that is completely unique from all other customers at AWS. So I'm just going to use numbers. I have to choose where I'm creating that actual bucket and then I can go through and decide how I'm going to use it. Now note, the default of creating a bucket is that nobody gets to actually access the bucket except the owner that's creating the bucket, that's me right now. I can change that, but that's on me. Once I go through all the steps and decide do I want to create versioning to have copies of my files just in case somebody deletes one. There's many, many, many different options. Would I like it to be encrypted? And I'll create the bucket, and once I create this bucket, good news, it was a unique, actual name. I can then go into the buckets properties and go further. Managing this bucket, what do I want to do? Do I want to set up rules to ensure that if content stay in this bucket longer than a certain period of time they're deleted, or maybe they're moved to the archive storage, S3 Glacier. Maybe I want to set up replication rules to replicate contents from one bucket to another bucket. So there's lots of different options to consider when creating a bucket as to how you want it to work and the level of durability. The other thing we can do with a bucket is actually go to the properties and take a look at how we define what happens when some changes occur to the bucket. So we can set up event notifications. If I create an event notification, I can give it a name maybe it's file upload, and then I can go through and decide what events I'm interested in. So somebody is putting something in my bucket. They're uploading, okay. That's what I want to hear about but how do I want to hear about it? After all the options we can go through and say, well, you could hear about it from a notification topic. So if you subscribed to this topic any time a file is uploaded to the bucket you'll hear about it. Maybe you want Lambda to take care of it and run some automation. So interesting options for your content yet again. The other options that we have available for storage are the shared storage options. Whether we want to pick EFS or FSX. We also could decide that we want to set up the storage gateway service and have content automatically moved from on-prem up into the cloud. So lots of different options. Let's sneak into EFS. If I create a file system, have to give it a name, I have to give it a network where it's going to be placed and I create it. Notice there's no drives. I don't see any of that. I've just got a certain amount of unlimited storage. Now, if I go into the file system then I can start actually looking at what's going on with the product and I can edit how it's going to run if I want to change the performance mode, and I can do this any time. Do I want it to burst? Do I want a certain amount of provision throughput? So I don't have EBS drives to tune, I have this storage array to tune, and there's lots of different settings that I can set up on the fly. So again, when I'm setting up some storage scenarios at AWS I don't get to see the drives. They're there but I don't interface with them directly. So setting up storage, there's a lot of different steps to consider. There's a quick overview of the different options that are available and we also could go into databases which are using those storage components as well. So spend some time with the storage options what's available at AWS and note you can change your decision that you make later on.