ESPN case
1. How many requests from users does
ESPN receive each second? Do You think this is a consistent pattern or does it
have peaks and valleys? When does it peak?
ESPN literally gets over 10.000 requests every second relentlessly over a timespan of hours during events. So this means that the peaks will be had during the events of which the speaker speaks (this means that the valleys will be when there are no events going on). These couple of hours there is a consistent stream of requests. I’m guessing during certain special events (such as the Super Bowl and the NBA finals) there will be an even higher amount of requests.
2. Why does ESPN store personal information and preferences on its databases and how does this personal information complicate the ability of ESPN to respond to requests from users? Why can’t ESPN just use Web page caching to handle the loads?
ESPN offers personalized content (this complicates the process because there is a need for an individual approach to every customer) therefor they need information from all individual requests and customers. This however increases the space that needs to be kept (on average roughly 5 kilobytes of data), that means there is a necessity for over 200 gigs of data/information which simply can’t be held by a reasonable number of Java Virtual Machines.
3. How much information on users does ESPN store? Why does this pose a challenge for ESPN? Can’t they just use a standard 1 terabyte hard drive from a PC? Why can’t they use a single PC?
On average ESPN will store 5 kilobytes of data. This reaches peaks of about 200 kilobytes for some customers and much less for others. It all depends on the amount of personalization. This means that they need to store a total of more than 200 gigs of data, this can’t be held any reasonable amount of Java Virtual Machines (they would need over 100).
This can’t be held on a simple PC or single hard drive because there is a necessity to code the information, so there is need for a program that can transform byte codes into computer language so it can be processed easily and faster.
4. What platforms do ESPN customers use when they access their Web sites, and how does this further complicate ESPN’s processing picture?
ESPN uses the same personalization system for both the computer based website as their mobile platform and subsequently for 2 other websites (soccernet and deportes) that ESPN runs. They do this because they see this as a part of the personalization process. They want the personalized experience (team preference) to be used throughout all their systems and on any device with which you can access the ESPN sites/apps.
Because all these different devices (tablets, computers, smartphones,…) need to be remembered and synchronized with the same preferences this will take up even more dataspace. But there will also be a need to adjust the information throughout the different devices to be more user-friendly and accessible.
5. What are the key components in ESPN’s solution? Describe the function of each.
Personalization DB is the system that ESPN uses, this consists of 2 specific components. Namely the Grid and the Composer. The Grid is effectively a component that encompasses the specific personalization inside the DB. This makes for a reaction time within a couple of milliseconds, The Grid makes for an ultrafast connection time to the personalization DB.
The second part if the Composer, this includes the differentiation between all the services that ESPN offers, and these are legion. As we saw before they take care of multiple sites and do all this on many different types of devices.
The normal process is: a request comes in and gets queued up in the composer, the composer goes to the Grid and get the preferences/personalization and within milliseconds your preferences will be sent back to you and you will get your preferred services.
6. Why is scalability so important to ESPN?
They need to be able to scale their service to ensure a fast response time when a request comes in. They do this by having 10 servers that they can operate at any given time. This system has an ability to dynamically spin down servers adjusting themselves to the necessity of the amount of requests. If less servers are needed, less will be used than they originally predicted. This coupled with the speedy response system makes for a good customer experience where downtime or slow reacting is almost never an issue. This system allows for many different personalized requests to come in but still be handled with relative ease.
ESPN literally gets over 10.000 requests every second relentlessly over a timespan of hours during events. So this means that the peaks will be had during the events of which the speaker speaks (this means that the valleys will be when there are no events going on). These couple of hours there is a consistent stream of requests. I’m guessing during certain special events (such as the Super Bowl and the NBA finals) there will be an even higher amount of requests.
2. Why does ESPN store personal information and preferences on its databases and how does this personal information complicate the ability of ESPN to respond to requests from users? Why can’t ESPN just use Web page caching to handle the loads?
ESPN offers personalized content (this complicates the process because there is a need for an individual approach to every customer) therefor they need information from all individual requests and customers. This however increases the space that needs to be kept (on average roughly 5 kilobytes of data), that means there is a necessity for over 200 gigs of data/information which simply can’t be held by a reasonable number of Java Virtual Machines.
3. How much information on users does ESPN store? Why does this pose a challenge for ESPN? Can’t they just use a standard 1 terabyte hard drive from a PC? Why can’t they use a single PC?
On average ESPN will store 5 kilobytes of data. This reaches peaks of about 200 kilobytes for some customers and much less for others. It all depends on the amount of personalization. This means that they need to store a total of more than 200 gigs of data, this can’t be held any reasonable amount of Java Virtual Machines (they would need over 100).
This can’t be held on a simple PC or single hard drive because there is a necessity to code the information, so there is need for a program that can transform byte codes into computer language so it can be processed easily and faster.
4. What platforms do ESPN customers use when they access their Web sites, and how does this further complicate ESPN’s processing picture?
ESPN uses the same personalization system for both the computer based website as their mobile platform and subsequently for 2 other websites (soccernet and deportes) that ESPN runs. They do this because they see this as a part of the personalization process. They want the personalized experience (team preference) to be used throughout all their systems and on any device with which you can access the ESPN sites/apps.
Because all these different devices (tablets, computers, smartphones,…) need to be remembered and synchronized with the same preferences this will take up even more dataspace. But there will also be a need to adjust the information throughout the different devices to be more user-friendly and accessible.
5. What are the key components in ESPN’s solution? Describe the function of each.
Personalization DB is the system that ESPN uses, this consists of 2 specific components. Namely the Grid and the Composer. The Grid is effectively a component that encompasses the specific personalization inside the DB. This makes for a reaction time within a couple of milliseconds, The Grid makes for an ultrafast connection time to the personalization DB.
The second part if the Composer, this includes the differentiation between all the services that ESPN offers, and these are legion. As we saw before they take care of multiple sites and do all this on many different types of devices.
The normal process is: a request comes in and gets queued up in the composer, the composer goes to the Grid and get the preferences/personalization and within milliseconds your preferences will be sent back to you and you will get your preferred services.
6. Why is scalability so important to ESPN?
They need to be able to scale their service to ensure a fast response time when a request comes in. They do this by having 10 servers that they can operate at any given time. This system has an ability to dynamically spin down servers adjusting themselves to the necessity of the amount of requests. If less servers are needed, less will be used than they originally predicted. This coupled with the speedy response system makes for a good customer experience where downtime or slow reacting is almost never an issue. This system allows for many different personalized requests to come in but still be handled with relative ease.