Chapter 9 checkpoint

Chapter 9 checkpoint: Pastebin and Bitly

Every question from this chapter’s lessons. Your first answer to each is the one that counts; afterward you can practice again as often as you like. Passing needs 75% right on the first attempt.

0 of 17 answered0 right on the first attempt

Question 1 of 17

Using the primer’s rule of thumb of 2.5 million seconds in a month, how many paste reads per second does 100 million reads a month average out to?

Your answerreads/s

Answers within 5% count.

Question 2 of 17

Each paste takes about 1.27 KB (content plus its database row), and 10 million new pastes arrive each month. How much new data accumulates in 3 years, in GB?

Your answer

Answers within 5% count, in any unit you choose.

Question 3 of 17

After 3 years there are 360 million shortlinks. You cache the most popular 20%, at about 500 bytes per cached entry. How much RAM is that, in GB?

Your answer

Answers within 5% count, in any unit you choose.

Question 4 of 17

A POST creates a new paste. Which response does RFC 9110 make the standard for a newly created resource?

200 OK with no header202 Accepted302 Found201 Created with a Location header

Question 5 of 17

How many different 7-character Base62 keys are there, in trillions?

Your answer

Answers within 2% count, in any unit you choose.

Question 6 of 17

With 360 million random 7-character Base62 keys already in use, a new random key collides with one of them about one time in how many?

Your answertries

Answers within 5% count.

Question 7 of 17

You generate shortlinks by Base62-encoding an auto-increment counter. What is the main drawback for unlisted pastes?

Encoding a number in Base62 is slowKeys are always longer than 7 charactersThe keys are guessable, so anyone can enumerate every pasteKeys collide often

Question 8 of 17

An application server takes a batch of keys from a key generation service and then crashes. What happens to the keys it had not used?

They are lost: wasted, but never handed out againThe KGS goes down until the server restartsThey return to the pool for other serversAnother server may get the same keys

Question 9 of 17

Shortlinks are base62(md5(ip_address + timestamp))[:7]. Two pastes from the same office, sent in the same second, keep getting the same shortlink. Why, and what fixes it?

Seven characters are too fewSame IP and timestamp give the same hash; hash random data insteadA cache is returning the first pasteMD5 is not uniformly distributed

Question 10 of 17

Why does this design keep paste content in an object store such as S3 and only metadata in the database?

Rows stay small so the index fits in memory, and the store grows cheaplyThe database cannot store textObject stores delete expired pastes automaticallyObject stores support richer transactions

Question 11 of 17

A paste expired an hour ago, but the hourly sweeper has not run yet, and a user requests it. What should happen?

Return 500 Internal Server ErrorThe read checks the expiry, returns an error such as 410 Gone, and triggers deletionServe it: it still exists until the sweeper runsRun the full sweeper first

Question 12 of 17

Match each expiry mechanism to how it behaves.

Redis, passive expiryRedis, active expiryDynamoDB TTL

Question 13 of 17

The sweeper must find expired pastes without scanning the whole table. What should the table store?

Nothing extra; the primary key is enoughAn indexed expires_at timestampOnly the expiration length in minutesA boolean expired column

Question 14 of 17

A shortener must count every click and let owners change where a link points. Which redirect should it send?

200 OK with the target page’s content308 Permanent Redirect302 Found or 307301 Moved Permanently

Question 15 of 17

Match each redirect status to its behavior.

301308302307

Question 16 of 17

Redirects are sent with Cache-Control: public, max-age=60 behind a CDN. The click counts in your server logs are far lower than the CDN’s request counts. Why?

A 302 can never be cachedThe CDN answers cached redirects itself, so those clicks appear only in its logsYour server logs are written late, so they miss recent clicksYour servers are dropping requests

Question 17 of 17

Analytics must now be near real time. Where should the click event be published?

From the browser cacheTo a log such as Kafka, after the redirect has been sentSynchronously to the analytics database, before redirectingNowhere; run the monthly MapReduce job more often