Have you noticed how slow the product details page is lately?

Conversation

Dialog

Roger: Hey Smith, have you noticed how slow the product details page is lately?

Smith: Yeah, I did. I checked the logs, looks like we’re hitting the database for every request. It’s a pretty expensive operation.

Roger: Exactly. I think we should introduce some caching to reduce latency.

Smith: Agreed. Maybe we can cache the data fragment for product specs. It rarely changes.

Roger: Good idea. Should we go with a cache-aside strategy?

Smith: That works. We can check the cache first and only hit the DB if it’s missing.

Roger: And to prevent stale data, maybe set a TTL of 5 minutes?

Smith: Sounds good. We could also invalidate the cache manually when a product is updated.

Roger: Right. Also, some of those price calculations involve a slow function call, we should cache that too.

Smith: Yep. Let’s add Redis into the mix and track the cache keys by product ID.

Roger: Perfect. I’ll write the caching layer and set up the fallback to DB.

Smith: Awesome. Let’s deploy it to staging and test the load time.

---

Vocabulary

🔹 hit (the database)
To make a request to the database to retrieve or update data.
Example: "We hit the database every time a user loads the page."

🔹 data fragment
A small piece or section of data, often used to refer to part of a larger dataset.
Example: "We only need to cache the data fragment for the product description."

🔹 have you noticed...?
A phrase used to bring attention to something that may have changed or needs observation.
Example: "Have you noticed how slow the page has become?"

🔹 stale data
Outdated information stored in cache that may no longer reflect the current state.
Example: "We need to refresh the cache regularly to avoid serving stale data."

🔹 reduce latency
To decrease the delay between a request and its response.
Example: "Caching will help reduce latency and improve user experience."

🔹 set a TTL of
To assign a Time-To-Live (TTL), meaning how long a cached item stays valid.
Example: "We can set a TTL of 5 minutes for product info."

🔹 slow function call
A function that takes a long time to execute, often due to complex logic or external dependencies.
Example: "That price calculation is a slow function call — let's cache the result."

🔹 expensive operation
An operation that consumes a lot of time, memory, or processing power.
Example: "Fetching the entire product list is an expensive operation."

🔹 track the cache keys
To monitor and manage the identifiers used to store and retrieve cached items.
Example: "We should track the cache keys using the product ID."

🔹 load time
The amount of time it takes for a web page or resource to fully appear for the user.
Example: "We want to improve the load time of the product details page."

🔹 (product) specs
Short for specifications — technical or descriptive details of a product.
Example: "The product specs include dimensions, weight, and materials."

🔹 (something) missing
When something expected is not present or found.
Example: "If the data is missing in the cache, we’ll fetch it from the database."

---

Exercises

Vocabulary

Choose the best option that matches the meaning or usage of each expression.

---

1. What does it usually mean when you “hit the database”?
A) Back up the database manually
B) Update the user interface
C) Make a request to the database
D) Delete a record permanently

2. What is a “data fragment”?
A) A file stored in the recycle bin
B) A piece of a larger set of data
C) A network protocol
D) A type of error message

3. “Have you noticed...?” is a way to:
A) Ask for permission
B) Report a technical failure
C) Request a password change
D) Bring attention to a change or issue

4. What is “stale data”?
A) Fresh data from the API
B) Data that is no longer valid or updated
C) Data stored in a CSV file
D) Data still loading

5. To “reduce latency” means to:
A) Add more logging
B) Decrease the response time
C) Increase CPU usage
D) Remove all caching

6. What does “set a TTL of” 10 minutes mean?
A) Set a timer to delete the database
B) Configure how long the cache remains valid
C) Trigger a file transfer
D) Make the page load slower

7. A “slow function call” typically:
A) Uses less memory
B) Is unrelated to performance
C) Takes a long time to execute
D) Is only used in testing

8. An “expensive operation” in programming usually involves:
A) High cost in time or system resources
B) External user authentication
C) Currency conversion logic
D) Buying additional software

9. To “track the cache keys” means to:
A) Encrypt cache data
B) Monitor and manage how data is cached and retrieved
C) Generate database indexes
D) Secure passwords

10. What does “load time” refer to in a web app?
A) Server downtime
B) Time it takes for a page or component to fully appear
C) Memory usage of a background job
D) Frequency of user sign-ins

11. “Product specs” typically include:
A) Customer email addresses
B) Marketing slogans
C) Technical details about a product
D) Billing rates

12. When “something is missing” in a cache layer, it means:
A) The user’s session expired
B) The item was not found in the expected storage
C) A full database reset occurred
D) The server restarted unexpectedly

---

Fill-in-the blanks

Complete each sentence with the correct word or expression from the list below:

Word Bank:
hit the database • data fragment • have you noticed • stale data • reduce latency • set a TTL of • slow function call • expensive operation • track the cache keys • load time • product specs • missing

---

1. To avoid unnecessary queries, the system first checks the cache before it needs to __________.

2. If you only need the title and price, cache that small __________ instead of the whole product object.

3. __________ how long the homepage takes to display lately?

4. We need to refresh the cache often, or users will see __________ that’s no longer accurate.

5. Using a CDN can help us __________ for users far from the server.

6. Let’s __________ 15 minutes so the cache doesn’t grow too stale.

7. That tax calculation runs every time and takes 2 seconds — definitely a __________.

8. Pulling all records with joins on 5 tables? That’s an __________ we should avoid frequently.

9. We should __________ by product ID to easily invalidate or refresh specific entries.

10. The __________ of our blog page increased after we added more third-party scripts.

11. The engineer asked for the __________ so she could double-check dimensions and compatibility.

12. If the data is __________ in the cache, we’ll fetch it directly from the backend.

---

Reading comprehension

1. Why are Roger and Smith discussing caching strategies?
A) They want to upgrade the server
B) The product details page is loading slowly
C) They need to store user passwords
D) They’re planning to delete the product table

2. What type of data do they decide to cache first?
A) User profiles
B) Search queries
C) Product specs
D) Shopping cart contents

3. What strategy do they plan to use for caching?
A) Full-page caching
B) Write-through caching
C) Cache-aside (lazy loading)
D) No-cache strategy

4. How do they plan to deal with stale data?
A) Never cache anything
B) Use a CDN
C) Manually update the cache and set a TTL
D) Disable database access

5. Which tool do they mention for implementing the caching layer?
A) PostgreSQL
B) Redis
C) MongoDB
D) Elasticsearch

---

Homework

Role play

Instructions:
Work with a partner and create a short dialog (10–14 lines) between two software engineers discussing how to improve the performance of an API or web page. Be sure to include at least 8 of the following words or phrases from the Word Bank in your conversation.

Word Bank: hit the database • data fragment • have you noticed • stale data • reduce latency • set a TTL of • slow function call • expensive operation • track the cache keys • load time • product specs • missing

---

Life experience

Reflect on your own experiences in software development or system design. Answer the following questions based on situations similar to the dialog and vocabulary you’ve learned:

1. Have you ever worked on a system or application where performance was an issue?
What was causing the slow load time, and how did you or your team try to reduce latency?

2. Can you recall a time when using cached data caused problems due to it being stale?
How did you detect it, and what did you do to prevent it from happening again?

3. Have you ever used tools like Redis or Memcached to avoid hitting the database repeatedly?
What kind of data or function calls did you cache, and what was the impact on performance?

4. In your projects, have you ever needed to track cache keys or set a TTL for cached items?
Why was this important in your system, and how did you manage invalidation or expiration?

---

Done

---