Guard caches against stampede

Prompt that protects caches from concurrent recomputation with coalescing and staggered expiry. Paste it into an AI tool of your choice, which then guards the cache against stampede.

Topic: Apps & prototypes
Type: Prompt
Level: Intermediate
Time: about 15 min

Results depend on the tool, the model and what you enter.

Fill in

What you type stays in this browser window. We store and send nothing.

CacheWhich cache?

OriginWhat fills the cache?

SymptomWhat do you see?

Your prompt

Help me protect a cache so an expired entry is not recomputed by many requests at the same time.
Cache: [Cache]
Origin: [Origin]
Symptom: [Symptom]
Steps: 1. Read the cache keys, expiry settings and the code that calls the origin on a miss. 2. Check whether each miss recomputes on its own or concurrent requests are coalesced. 3. Propose a short lock or coalescing per key, with expiry if the computing run crashes. 4. Decide whether stale entries may still be served briefly while a refresh runs in the background. 5. Check whether all entries expire in the same second and propose staggered expiry. 6. Name metrics and alerts for hit rate, origin load and wait time.
Output: current state, coalescing design, rule for stale entries, expiry proposal and alerts.

Take particular care to coalesce concurrent recomputations and do not let the origin collapse under an empty cache.

Before you answer, check: Are all instructions met? If key information is missing or a placeholder in square brackets is still in the text, ask instead of guessing. Mark assumptions as assumptions.

Help me protect a cache so an expired entry is not recomputed by many requests at the same time.
Cache: [Cache]
Origin: [Origin]
Symptom: [Symptom]
Steps: 1. Read the cache keys, expiry settings and the code that calls the origin on a miss. 2. Check whether each miss recomputes on its own or concurrent requests are coalesced. 3. Propose a short lock or coalescing per key, with expiry if the computing run crashes. 4. Decide whether stale entries may still be served briefly while a refresh runs in the background. 5. Check whether all entries expire in the same second and propose staggered expiry. 6. Name metrics and alerts for hit rate, origin load and wait time.
Output: current state, coalescing design, rule for stale entries, expiry proposal and alerts.

Take particular care to coalesce concurrent recomputations and do not let the origin collapse under an empty cache.

Before you answer, check: Are all instructions met? If key information is missing or a placeholder in square brackets is still in the text, ask instead of guessing. Mark assumptions as assumptions.

This is an instruction for an AI, not the result itself.

Where this prompt works

These AI tools are a good fit for this prompt:

How to use it

Fill in the fields or tap a suggestion. Leave out names, addresses and other personal data.
Copy the finished prompt into an AI tool of your choice.
Paste cache configuration, the miss-path code and metrics for hit rate and origin load, without real customer data.

Why does this work?

When a popular entry expires, hundreds of requests often hit the database at once. Without coalescing, load multiplies exactly when the cache is empty.

Tips

Test coalescing and briefly serving stale values under artificial load on a test system first.

You may use our prompts freely, including at work. Please do not republish them as your own collection. We give no warranty for AI results; check them yourself. Terms of use

Guard caches against stampede

Fill in

Your prompt

Where this prompt works

How to use it

Why does this work?

Tips

Also a good fit

Choose pagination for an API list New

Write an API reference people can use right away New

Plan a throttled data backfill New