Amazon AWS Certified Advanced Networking Specialty – Billing Section

1. Direct Connect Billing

Hi everyone and welcome back. In today’s video we will be discussing about the billing aspect as far as the Direct Connect is concerned. So till now we have been discussing in great detail related to Direct Connect and some of the technical aspects. However, it is also important to understand how much Direct Connect will cost you in the organization. So let’s go ahead and understand more about it. Now, before we go ahead and understand the Direct Connect billing, let’s look into the typical AWS billing and how generally AWS charges for its resources. So generally there are a few factors based on which the AWS charges. First is typically for services like EC two, you get charged per hour or per second. Billing is also introduced. You have the network pricing and you also have the resource charges. So if you take an example of an S Three bucket, so in S Three bucket, the first thing that you will be charged is the storage, the amount of storage that you utilize.

Second is the network pricing. So what is the data transfer in and what is the data transfer out? And third is the resource charges. How many million write requests, how many million get requests you are making to the S Three bucket is also charged. Now, when it comes to the network pricing, we have already seen that you have a data transfer in and data transfer out. So data transfer in is completely free. AWS will not charge you for that. However, the data transfer out charges are something that you will be charged for.

So let’s assume that you have put a video file in the S Three bucket. Now, I request for that video file, so I send an Http get request for that video file. So that request which is the data transfer in would be free, but the data transfer out the actual video file which is being streamed from the S Three bucket to my browser, that is part of the data transfer out and that is something that will be charged. So now coming to the Direct Connect billing, indirect connect, the bill is actually calculated based on two factors. One is pricing per port hour. This is first and the data transfer charges. Again, the data transfer in is free and the out is charged. So many might assume that Direct Connect might be pretty expensive to set up in AWS, and actually it is not.

So Direct Connect is completely free for setup. If you’re going with a partner base, then there might be certain charges they charge based on cross connect. So there might be different charges. But if you go with AWS, the setup is free, but you will be charged based on port hour and you will be charged on data transfer. Now, one important part to remember over here is that account owner where the resource is present pays the price. This is very important pointer to understand. So let’s take an example on what I mean by this. So, when it comes to virtual interfaces. Now, virtual interfaces, if you remember that we have to accept the virtual interface connection from the Direct Connect console. So the account, this specific virtual interface can be hosted on your account, or it can even be hosted on the external account.

Now, if the virtual interface is hosted on an external partner account, then the partner will have to pay for the data transfer within this virtual interface. All right? So this is something that you need to remember. So if you’ll see over here, I have a customer account A, where I have a public wif now, before this public with can actually be functioning, you have to accept the viv connection from the Direct Connect console. So that except the account from which you accept that VIFF connection has to pay for the price of the data transfer associated with this VIP. So, quite important to understand, just remember this because it might be a little confusing and in exam you might get confused saying that if both the Whiff belongs to account A and you have hosted that one of the wife in an external account B, who will be paying the pricing. And the answer is the pricing will be paid from the account where this WIFF is hosted. All right? So one more important part to understand is that in US.

There is a separate rate applied when you access the interregion resources. So Direct Connect does have the capability to access the interregion resources in the US region and there are certain separate charges that are being applied. Now, one important part that I would like to stress upon is that do read the FAQs of Direct Connect and within the FAQs you will get actually a lot of clarity upon the pricing which is involved. And the first thing if you will see there are no really set up charges for Direct Connect. So you get billed per port hour so again if you go into the Direct Connect. So this is the plan. So this is port speed of 5100 201 gig. And ten gig. So 50 is 0. 3 /hour so this is the port hour speed. For all the regions except Japan however, if you go with other ones like ten gig, you have to pay for $2. 25 per hour for port and along with that you have to pay for the data transfer out. So if you’ll see over here the data transfer in, data transfer in is 0. 0 per GB in all locations. So this is free. However, the data transfer out is something that you have to pay for.

2. Data Transfer charges

Hey everyone and welcome back. In today’s video, we’ll be discussing about the data transfer charges in AWS. Now, AWS data transfer costs are basically the cost which is associated with transferring the data, which can either be within AWS environment, like between various AWS services like EC Two to S Three, etc. Or it can be between AWS and public Internet. So generally, the data transfer cost is categorized into four major sections. First is Internet, then is Region to region. Then you have inter availability zone. And the fourth one is VPC peering. So at a high level overview, it is important for us to understand the costs associated with all four of these categories. So, let’s discuss about the data transfer costs associated with the internet. Now, if the data is going outside of the AWS towards the public Internet, it will be charged at $0. 9 per GB up to the first ten TB. Now, here one important part to remember is this public Internet. Now, the public internet is equal to from AWS public IP to a non AWS public IP.

This is important part to remember because lot of AWS services are global. And if you are sending data to those AWS services which has a global IP, that is not considered as a public Internet, public Internet is only considered when the data has been sent from a public IP to a non AWS public IP. All right. So same thing is explained in the third point where the above definition. So this first definition excludes the traffic between two AWS regions or traffic between public IP of the same AWS region. Now, along with that, one important part to remember is that there is no data transfer charge between EC Two and other Amazon services within the same region. For example, data transfer charge between EC Two in the US West region and the S three in the US West region. Now, the same can be better explained with a diagrammatic representation here where you have the EC Two instance which is running in a Mumbai region and it is sending the data out towards the Internet. So with this, I hope you understood what the internet means in this type of use case.

So, Internet basically means that any destination which has a non EWS public IP. So the data transfer cost is $0. 9 per GB and this is the data transfer out cost. So the data transfer in cost is zero. You are not charged for that. So this is the reason why any data that you send towards the Internet will be charged at a cost of $0. 9 per GV. However, whatever data that comes inside is free of cost. Now, let’s look into the data transfer cost at the regional level. So, first is the data transfer cost within the Availability Zone. Now, the data transfer cost in the same region and within the same Availability Zone is zero. With one condition that you must be using a private IP address.

So let’s say that you have two easy two instances and you are sharing a lot of data between those two EC two instances. Now, if you’re sharing the data between the private IP addresses, then you will not get charged. If you are doing it via the public IP address, you will get charged. All right? So this is within the same Availability Zone. Now let’s discuss the charges which are occurred when the data is moved across Availability Zone within the same region. Now, the data transferred between AWS services which are located in the same region, but in the different Availability Zone is considered as the regional data transfer and is charged at $0. 1 per GB for outgoing data transfer. Now, along with that, if you are having a VPC Peering, for VPC Peering, also you will be charged at $0. 1 per GB.

So when you talk about the cost of an interregion, so this is the diagrammatic representation of an interregion where you are transferring between the regions. So this is costed at $0. 2 per GB. So, understanding at a high level overview about the things that we discuss, so, we already discussed that if the EC two instances are within the same Availability Zone, you will not get charged for the data transfer. If the instances are within the different Availability Zone, then you will be charged at $0. 1 per GB. If you are sending the data between regions, you will be charged at $0. 2 per GB. Now, for the exams, you don’t really need to remember the exact costing here, because this costing is subjected to change. But at a high level overview, you should be able to understand on how the cost thing is done in AWS with respect to data transfer.

So, before we conclude, let me quickly show you a few more things. So, you have a simple monthly calculator over here, and from this calculator you can figure out the cost. So this is one of the easiest way. So, let’s say we are putting VPC Peering data transfer and this is charges $0. 1 per GB. Now, one important part to remember here is that do not take this cost as a literal term, because this is subjected to change. So AWS might revise it anytime.

So you don’t really need to remember the cost, but you need to remember at what stage and what layer the data transfer costs apply. So even in the EC, to instance pricing, if you go a bit down, it basically shows you the data transfer cost. So if you see the data transfer in to EC Two from the Internet is zero. So you are not charged for that. So, data transfer out from EC Two to Internet for up to one GB per month is not charged. But above one GB, it is charged at $0. 9 per GB. All right? So you should be able to understand at what layer the transfer costs are being applied. And that is the prime motive about the lecture.

3. ELB Billing

Hey everyone and welcome back. In today’s video we will be discussing about the ELB billing. Now, in AWS Elastic Load Balancer service is offered in terms of three categories. First is application load balancer, second is network load balancer and third is classic load balancer. Now, one very important part to remember over here is that AWS will charge you for the ELB even if there are no so easy to instance attached to it. So if you have idle ELB somewhere in your account, you will be charged for that. I still remember that one of my colleagues, he was using Free Tire account and after his Free Tire account, he had just created an ELB for testing purpose and he forgot to delete it. Now at the end of that he got a bill of 20 USD. Now he was asking, hey, I don’t really have any EC two instance attached to the ELB, why am I getting charged? And basically he should have studied the ELV billing for that. Anyways, so this is one important part to remember.

Now discussing about the ELV billing, the way the billing happens in the new generation of load balancer, which is application load balancer, network load balancer is little different when you compare with the classic load balancer where billing was pretty simple. So in terms of application load balancer, you basically get charged for each hour or partial hour that an application load balancer is running for and the number of load balancer capacity units used per hour. So there are two things where you get billed. First is the number of R your load balancer is up and running. And the second thing is the LCU Warrior. So LCU is pretty important. Now, similar goes for the network load balancer. You get charged for each hour or partial R that the network load balancer is running and the number of LCU used by the network load balancer.

So the first is you get charged per R or partial R and the number of LCU. So for application and network load balancer you have the running time cost and the LCU. In terms of classic load balancer, you do get charged for R or partial R that the classic load balancer is running for and for the number of GB of data transferred through your load balancer. So this is specifically for the classic load balancer. Now, when we were discussing about the application load balancer and the network load balancer here, the LCU part is important to understand because understanding that, all right, application load balancer has ran for 10 hours, so you pay for 10 hours. So that is pretty simple to understand. But the LCU is something that we need to look into. So this is a jam-packed slide over here, but let’s keep it simple.

So basically, LCU is divided into four dimensions. One is new connection, active connection, processed bytes also can be referred as bandwidth and the rule evaluation. Now, rule evaluation is only for the ALB, because in ALB you can have multiple rules. For network load balancer. The last point of rule evaluation does not really come into picture. So these are the metrics based on which an LCU offer application load balancer is counted into. For network load balancer, rule evaluation part is no longer applicable. So new connection is basically number of newly established connections per second. Pretty simple to understand.

Active connection is number of active connections per minute. So if you have a TCP connection which is open and established, that can be considered as an active connection, then you have the processed byte and you also have the rule evaluation. Now, specifically for rule evaluation, we know that it’s applicable for ALB. It’s important to remember that if you have ten or fewer rules configured, then the rule evaluation dimension is ignored in the LCU computation. Now, among all of these four factors, you are charged only on the dimension which has the highest usage.

So your billing will look into each of these dimensions, it will look into which of this dimension has the highest LCU and you will only be charged based on the highest LCU dimension. So let’s go ahead and understand it in a much more better way. So, one LCU basically contains 25 new connections per second, 3000 active connections per second, 1000 rule evaluations per second, as well as one GB per hour for EC, two instance and . 4 GB per hour for the lambda function as a target. Now, by this it you will not come to know how exactly the billing would work. So we have a next slide for that. So this slide is based on a use case where let’s say your organization is receiving one connection per second.

So you are receiving one new connection per second. So in order to calculate the LCU, what you do is you divide one by 25. So 25, basically this is your you can say this is the static. So one LCU contains 25 new connections per second, but you are just getting one connection per second. So you divide one by 25. So your LCU here is 0. 4. Now, for active connections, you are receiving 100 active connections per minute. Now, if you look into the LCU, it’s 3000 active connections per minute. So you divide 100 by 3000 and you get 0. 3 LCU here. All right? Now processed bytes, processed bytes is one GB per R for EC two instance. Now you are getting two GB per R of data transfer. So here your LCO becomes two and rules. You have currently ten processed rules.

So this is not applicable over here. So this is a use case for the application load balancer. If you are having network load balancer, the rule portion does not apply at all. So among these three, you notice which one is the highest. The process byte LCU is the highest, which is two LCU. So now you will be charged based on two LCU. So now based on the cost of one LCU, let’s calculate. So for ALB one LCU costs $0. 8. Now, the maximum LCU that we have is two. So the LCU associated with new connection and active connection will be ignored and only you will be taking the highest LCU metric. So our highest LCU is two. So now you multiply two LCU into 0. 8, that comes to zero point $16. Now you multiply it into 24 US. So if you want to see what would be the cost into 24 hours, which is zero point 38 four. Now you do into 30. So 30, if you multiply you get eleven point $52. So for this specific use case, you will be charged eleven point 52.

Now, important part to remember, we already discussed that ALB and NLB pricing are based on LCU first and also based on the running cost. So the second point is the running cost. So for ALB the cost of running per R is zero point 25. Now if you multiply it into 24, it comes to 0. 5. And then if you look into how much it will cost in 30 days, it will cost 16. 2. So you add 16. 2 plus eleven point 25, it comes to close to around 28 USD that will be charged based on the use case that you have. Now, do remember that this is an approximate pricing. AWS generally keeps on revising the cost. So that is one important part to remember.

Now, before we conclude, if you basically look into the elastic load balancer pricing, you have application load balancer, you have network load balancer. You have the classic load balancer. Now the application load balancer, if you look into the LCU it cost 0. 008. Now network load balance is much cheaper. It costs 0. 006 per LCU hour. So you have two costs. One is the running of the load balancer and second is the LCU. So this keeps on changing. So this is one important part to remember. Now, again, in examination you will not be asked to calculate the LCU and calculate the pricing. But you’ve, you should be aware about the high level overview on how exactly you will be charged on the load balancers.

• Category: Uncategorized