What Is a Cache Miss? Definition With ASP.NET Core Examples

If you search what is a cache miss, you want a plain definition before Redis tutorials or performance tuning. Here it is:

A cache miss occurs when a requested piece of data is not found in the cache (or has expired), so the application must load it from the slower backing store — usually SQL Server, an HTTP API, or disk.

The opposite is a cache hit: the value is already in memory or Redis and returns quickly.

Implementation: Redis caching in ASP.NET Core. Caching layers: caching system in .NET.

Cache hit vs cache miss

Request for key "tenant:1:providers:list:v1"

  CACHE HIT                          CACHE MISS
  ─────────                          ──────────
  Look up key → found                Look up key → not found (or expired)
       │                                    │
       Return cached JSON                   Run SQL / HTTP (slow path)
       (fast — microseconds)                      │
                                            Store in cache (optional)
                                                  │
                                            Return fresh data
                                            (slow — milliseconds+)

	Cache hit	Cache miss
Key in cache?	Yes, not expired	No, or TTL expired
Typical latency	Very low (RAM / Redis)	Full query or network round-trip
Load on SQL	None for that request	Full read (or write on rebuild)
User experience	Snappy dashboard	Slower; spikes if many miss at once

What is a cache miss? (definition)

Definition: A cache miss is a cache lookup that fails to return a valid entry, forcing the system to populate from the origin (database, file, remote service).

Miss types you will hear in interviews and ops:

1. Cold miss (compulsory) — first request ever for that key; cache is empty
2. Capacity miss — cache evicted the entry because memory is full
3. Conflict miss — poor key design causes unnecessary misses (e.g. missing tenant in key)
4. Expiration miss — TTL elapsed; entry removed intentionally

All four feel the same to the caller: miss → rebuild → (maybe) set.

Real-world analogy

A pharmacy shelf:

• Hit — the medicine is on the shelf; hand it to the customer
• Miss — shelf empty; walk to the stockroom, fetch a box, restock the shelf, then serve

If fifty customers ask for the same item and the shelf is empty, fifty trips to the stockroom happen unless one person restocks while others wait — that is cache stampede control (Redis guide).

Cache miss in ASP.NET Core

Typical IDistributedCache read path:

public async Task<ProviderListDto> GetProvidersAsync(ProviderQuery query, CancellationToken ct)
{
    var key = BuildKey(query);
    var cached = await _cache.GetStringAsync(key, ct);

    if (cached is not null)
    {
        // CACHE HIT — deserialize and return
        return JsonSerializer.Deserialize<ProviderListDto>(cached)!;
    }

    // CACHE MISS — expensive path
    var fresh = await _query.ExecuteAsync(query, ct);

    await _cache.SetStringAsync(
        key,
        JsonSerializer.Serialize(fresh),
        new DistributedCacheEntryOptions
        {
            AbsoluteExpirationRelativeToNow = TimeSpan.FromMinutes(5),
        },
        ct);

    return fresh;
}

Every miss runs _query.ExecuteAsync — often EF Core against SQL. If that query is slow (N+1, missing index), caching only helps after the first miss per key per TTL window.

Rule: tune SQL first (EF Core performance), then cache slim results.

Why cache misses spike in production

Event	What happens
Deploy / cold start	In-memory cache empty; Redis may be warm if shared
TTL expiry	Many keys expire together → synchronized misses
Traffic spike	Hot keys miss if never cached or evicted
Wrong key	Tenant missing → treat as miss every time or worse, serve wrong tenant’s hit
Cache flush	Ops clears Redis; all reads miss until rebuilt

Monitor hit ratio and miss rebuild duration. A “fast” API with 5% hit rate is not winning.

Reducing cache miss pain

1. Absolute TTL matched to freshness — not “cache forever”
2. Single-flight / lock on rebuild — one SQL query per key during stampede
3. Warm-up after deploy for known hot keys
4. Version in key (v1 → v2) instead of flushing entire prefixes blindly
5. Slim DTOs — a miss that serializes 2 MB to Redis hurts network too
6. Fail-open if Redis is down — miss every time but SQL still answers (connection errors)

Cache miss vs “slow SQL”

Caching hides slow SQL on hits. On misses, you still pay full query cost — sometimes plus cache write overhead.

Bad strategy:  N+1 query + Redis
               miss = slow every TTL window
               hit  = fast but stale wrong data risk

Good strategy: tuned SQL projection + Redis
               miss = acceptable once per TTL
               hit  = fast and correct enough

CPU cache miss (related term)

In computer architecture, CPU cache miss means the processor did not find data in L1/L2 cache and fetched from RAM — a different layer than application Redis. Interviewers may mention both; this article is application-level caching for APIs.

If an interviewer asks

"What is a cache miss?"
Lookup failed — data loaded from origin (SQL, HTTP) instead of cache.

"Hit ratio?"
hits / (hits + misses). Low ratio means cache is not helping or TTL is too short.

"Thundering herd on miss?"
Many concurrent requests miss the same key and all hit SQL. Fix: lock, single-flight, or early warm-up.

"When not to cache?"
Per-user PHI without strict keys, data that must be real-time, or SQL that is already broken.

More: Redis caching ASP.NET Core. Hub: caching.