Error Establishing a Redis Connection in ASP.NET Core (Fix Guide)

Error establishing a Redis connection is the message you see when StackExchange.Redis (behind IDistributedCache) cannot open a socket to the server — wrong host, TLS mismatch, firewall, or Redis simply not running. This guide walks through fixes on local Docker, Azure Cache for Redis, and ASP.NET Core configuration.

Caching concepts: what is a cache miss. Production cache design: Redis caching ASP.NET Core.

What the error looks like

Typical logs or exceptions:

StackExchange.Redis.RedisConnectionException:
  It was not possible to connect to the redis server(s).
  Error connecting to Redis.

WordPress / plugin messages (same root cause):
  Error establishing a Redis connection

ASP.NET Core surfaces this on first GetStringAsync / SetStringAsync if the multiplexer cannot connect.

Quick diagnosis checklist

□ Is Redis running? (docker ps / Azure portal status)
□ Host and port correct? (6379 plain, 6380 TLS on Azure)
□ Password / access key in connection string?
□ TLS enabled when server requires SSL?
□ From Docker Compose: hostname = service name, not localhost
□ Firewall / NSG allows outbound to Redis
□ Connection string in appsettings / Key Vault — no stale copy

Connection string examples

Local Docker Redis (no TLS)

{
  "Redis": {
    "Connection": "localhost:6379"
  }
}

builder.Services.AddStackExchangeRedisCache(options =>
{
    options.Configuration = builder.Configuration["Redis:Connection"];
    options.InstanceName = "myapp:";
});

Azure Cache for Redis (TLS required)

{
  "Redis": {
    "Connection": "yourcache.redis.cache.windows.net:6380,password=YOUR_ACCESS_KEY,ssl=True,abortConnect=False"
  }
}

Setting	Local	Azure
Port	6379	6380 (SSL)
SSL	false / omitted	ssl=True
Password	optional	access key required
abortConnect	optional	often False so app starts even if Redis briefly unavailable

Common mistake: using port 6379 against Azure without SSL — connection fails immediately.

Docker Compose: wrong hostname

From another container, localhost is the container itself, not the Redis service:

services:
  api:
    environment:
      - Redis__Connection=redis:6379   # service name
  redis:
    image: redis:7-alpine

Using localhost:6379 inside the API container → error establishing a Redis connection.

See also: Docker .NET + Angular local.

Test Redis before blaming ASP.NET Core

# Local
redis-cli -h localhost -p 6379 ping
# PONG

# Azure (TLS)
redis-cli -h yourcache.redis.cache.windows.net -p 6380 -a YOUR_KEY --tls ping

If redis-cli fails, fix infrastructure first — no C# change will help.

Azure App Service specific fixes

1. Connection string in Configuration → Application settings (not only appsettings.Production.json in repo)
2. Firewall — allow App Service outbound IPs or use VNet integration + private endpoint
3. TLS 1.2+ — enabled by default on modern stacks; very old clients rare on .NET 8+
4. Non-SSL port disabled — Azure often disables 6379; use 6380 + SSL

Deploy notes: Azure App Service ASP.NET Core.

Fail-open so Redis outages do not take the API down

If cache is an optimization, not a requirement:

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

    try
    {
        var cached = await _cache.GetStringAsync(key, ct);
        if (cached is not null)
            return JsonSerializer.Deserialize<ProviderListDto>(cached)!;
    }
    catch (RedisConnectionException ex)
    {
        _logger.LogWarning(ex, "Redis unavailable — cache miss path via SQL");
    }

    return await LoadFromSqlAsync(query, ct);
}

Every request becomes a cache miss while Redis is down — SQL must handle load. That is usually better than 500 on every read.

Timeouts and connection storms

yourhost:6380,password=...,ssl=True,abortConnect=False,connectTimeout=5000,syncTimeout=5000

• connectTimeout — how long to wait for initial connection
• abortConnect=False — allow app to start; retry in background (common on Azure)

Avoid creating a new ConnectionMultiplexer per request — register IDistributedCache once in DI.

When the error is not Redis at all

Symptom	Real cause
Works in Swagger, fails in Angular	CORS / wrong API URL — not Redis
Intermittent timeouts	SQL slow; mistaken for cache
After deploy only	Wrong slot settings / missing Key Vault reference
WordPress plugin message	PHP Redis extension — out of scope; same network fixes apply

Production checklist

1. Redis health in staging mirrors prod (TLS, firewall)
2. Connection string from secure config — never commit keys
3. Fail-open for optional read caches
4. Alert on Redis connection exceptions + miss-rate spike
5. Load test with Redis down — API should degrade, not die

If an interviewer asks

"Redis connection fails on startup?"
Check host, port, TLS, password; abortConnect=False for Azure; verify with redis-cli.

"Cache required vs optional?"
Read-through catalog → fail-open. Global rate limit counter in Redis → may need fallback policy or edge limiting.

Hub: caching.