Examples of Resource Estimation

Introduction#

Now that we’ve set the foundation for resource estimation, let’s make use of our background knowledge to estimate resources like servers, storage, and bandwidth. Below, we consider a scenario and a service, make assumptions, and based on those assumptions, we make estimations. Let’s jump right in!

Number of servers required#

Let’s make the following assumptions about a Twitter-like service.

Assumptions:

• There are 500 million (M) daily active users (DAU).
• A single user makes 20 requests per day on average.
• Recall that a single server can handle 8,000 RPS.

Indeed, the number above doesn’t seem right. If we only need 15 commodity servers to serve 500M daily users, then why do big services use millions of servers in a data center? The primary reason for this is that the RPS is not enough to estimate the number of servers required to provide a service. Also, we made some underlying assumptions in the calculations above. One of the assumptions was that a request is handled by one server only. In reality, requests go through to web servers that may interact with application servers that may also request data from storage servers. Each server may take a different amount of time to handle each request. Furthermore, each request may be handled differently depending upon the state of the data center, the application, and the request itself. Remember that we have a variety of servers for providing various services within a data center.

Therefore, we approximate the number of servers by depicting how many clients a server handles on a given day:

$$\frac{Number\ of\ daily\ active\ users}{RPS\ of\ a\ server}$$

In this case, it’s equal to: $\frac{500 M}{8000}$ = 62,500.

Note: This may not be an accurate estimation, but it’s a realistic one. Therefore, we use this approach in estimating the number of servers in our design problems. Informally, the equation given above assumes that one server can handle 8,000 users. We use this reference in the rest of the course as well.

• Total storage required for one day = $0.1875 TB + 15 TB + 112.5 TB \approx 128 TB$
• Storage required for one year = $365 \times 128 TB = 46.72 PB$

Bandwidth requirements#

In order to estimate the bandwidth requirements for a service, we use the following steps:

1. Estimate the daily amount of incoming data to the service.
2. Estimate the daily amount of outgoing data from the service.
3. Estimate the bandwidth in Gbps (gigabits per second) by dividing the incoming and outgoing data by the number of seconds in a day.

Incoming traffic: Let’s continue from our previous example of Twitter, which requires 128 TBs of storage each day. Therefore, the incoming traffic should support the following bandwidth per second:

$$\frac{128 \times 10^{12}}{86400} \times 8 \approx 12 Gbps$$

Note: We multiply by 8 in order to convert Bytes into bits because bandwidth is measured in bits per second.

Outgoing traffic: Assume that a single user views 50 tweets in a day. Considering the same ratio of five percent and 10 percent for videos and images, respectively, for the 50 tweets, 2.5 tweets will contain video content whereas five tweets will contain an image. Considering that there are 250 M active daily users, we come to the following estimations:

Twitter will need a total of $12\ Gbps$ of incoming traffic and $197.5\ Gbps$ of outgoing, assuming that the uploaded content is not compressed. Total bandwidth requirements = $12 + 197.5 = 209.5 \ Gbps$.