Caching & CDN Errors

Understanding Caching errors

Cache problems are rarely reported as cache problems. They arrive as a traffic spike that takes down the origin (a stampede after a mass eviction), users seeing old content after a deploy, or an inexplicable 403 from the CDN. The unifying diagnostic is to look at the cache status header on a real response before theorising about anything else.

How to debug Caching errors

1. Read the cache status header on a live request: curl -sI https://example.com | grep -i -E 'cache|age|x-cache|cf-cache-status'. HIT, MISS and BYPASS each point at a different root cause.
2. Check the Age header against your intended TTL. An Age larger than max-age means something is serving stale content deliberately (stale-while-revalidate), often the desired behaviour, occasionally the bug.
3. For stampedes, add request coalescing or a short randomised TTL jitter rather than a longer TTL. Identical expiry times across many keys are what create the thundering herd.
4. Use content-hashed filenames for static assets and no-cache for HTML. Almost every "users see the old version" incident traces back to a long TTL on an HTML document.
5. For CDN 403s, distinguish a CDN-generated response from an origin one by checking whether CDN-specific headers are present. An origin 403 has a completely different fix.

Tools worth reaching for

• curl -sI
• varnishlog
• CDN edge logs
• Cache-Control validators

All 10 Caching errors

• Browser serving stale cache after deploymentUsers see old content after a deployment because their browsers cached the previous version. Cache-busting is not…
• Cache stampede / thundering herdMultiple processes simultaneously attempt to rebuild the same expired cache entry, overwhelming the backend database…
• CDN: High cache miss ratioThe CDN is not caching responses effectively. Most requests are hitting the origin server, defeating the purpose of…
• CloudFront: 403 Access DeniedCloudFront cannot access the S3 origin, usually because the bucket policy or Origin Access Control (OAC) is…
• Memcached: SERVER_ERROR object too large for cache NewThe item exceeded the maximum item size, one megabyte by default, so it was refused rather than stored. Because most…
• Memcached: SERVER_ERROR out of memoryMemcached cannot allocate memory for new items. The memory limit has been reached and no items are eligible for…
• Redis: LOADING Redis is loading the dataset in memory NewThe server is replaying its RDB or AOF file after a restart and refuses commands until it finishes. On a large dataset…
• Redis: MISCONF Redis is configured to save RDB snapshots but is unable to persist NewA background save failed, usually because the disk is full, the dump directory is not writable, or the fork ran out of…
• Redis: NOSCRIPT No matching script. Please use EVAL. NewEVALSHA runs a script the server has already cached by hash, and that cache is not persisted. A restart, a failover to…
• Varnish: 503 Backend fetch failedVarnish cannot fetch content from the origin server. The backend may be down, too slow, or returning errors.

Other categories

• AI 35Rate limits, context windows, GPU memory and model-serving failures.
• Ansible 10Unreachable hosts, become passwords, undefined variables and Jinja2 failures.
• API 14Auth headers, payload limits, versioning, idempotency and webhook signatures.
• Apple 10Command line tools, dyld, Homebrew permissions, notarisation and Keychain.
• Auth 11OIDC, SAML, Auth0, Okta, Keycloak, passkeys and MFA failures.
• BigData 11Spark, Kafka, Airflow, Snowflake, Flink and Databricks failures.
• C# 12NuGet restore, null references, EF Core migrations, async deadlocks and Blazor…
• C++ 11Segfaults, linker errors, memory corruption and template deduction failures.
• CI/CD 18GitHub Actions, GitLab CI, Jenkins, CircleCI: permissions, runners and…
• Client 21CORS, mixed content, module resolution, memory limits and framework runtime…
• Cloud 25IAM permissions, quotas, service limits and credential failures.
• Dart 10Null safety, pub version solving and build toolchain problems.
• Database 41Connections, deadlocks, constraints, replication and memory limits.
• DNS 10NXDOMAIN, SERVFAIL, timeouts, propagation and delegation problems.
• Docker 27Daemon connectivity, disk space, image pulls, ports and architecture mismatches.
• Elixir 9GenServer timeouts, supervision failures and Mix compilation problems.
• Email 8Delivery failures, relay denial, authentication, SPF, DKIM and DMARC.
• Frontend 23Hydration mismatches, bundler resolution, layout shift and font loading.
• Git 20Merge conflicts, rejected pushes, detached HEAD, LFS and repository corruption.
• Go 19Nil map assignment, concurrent map access, context cancellation and deadlocks.
• GraphQL 13Validation, depth limits, N+1 queries and fragment problems.
• gRPC 10Status codes, deadlines, message limits, TLS and HTTP/2 transport failures.
• HTTP 494xx client errors, 5xx server errors, redirects, headers and protocol problems.
• ICMP 23Destination unreachable, time exceeded, fragmentation needed and redirects.
• Ingress 8404 default backend, missing TLS secrets, IngressClass and path matching.
• Java 19Class loading, dependency resolution, connection pools and JVM version…
• JavaScript 42npm resolution, async pitfalls, hydration, memory limits and runtime type…
• Kubernetes 34CrashLoopBackOff, ImagePullBackOff, OOMKilled, RBAC, scheduling and storage.
• Logging 9Log4j, Logback, Fluentd, Logstash and CloudWatch ingestion problems.
• MessageQueue 14Kafka, RabbitMQ, SQS, NATS and Celery: lag, rebalancing and poison messages.
• Mobile 17Gradle, CocoaPods, Xcode signing, Metro bundler and toolchain mismatches.
• Monitoring 12Prometheus scrapes, Grafana data sources, OpenTelemetry exporters and agent…
• Network 35Refused connections, timeouts, resets, MTU problems and port exhaustion.
• Performance 6GC pauses, thread pool starvation and event loop blocking.
• PHP 11Memory limits, execution timeouts, autoloading, Composer and PDO connections.
• Proxy 17nginx, Envoy, HAProxy, Traefik, Caddy and Cloudflare upstream failures.
• Python 35Imports, virtual environments, encoding, concurrency and dependency conflicts.
• Regex 8Catastrophic backtracking, back references, escaping and engine differences.
• Ruby 11Bundler, migrations, native extensions, encoding and asset compilation.
• Rust 19Borrow checker, ownership moves, trait bounds and lifetime mismatches.
• Scala 6Dependency resolution, binary compatibility and type inference failures.
• Security 25JWT validation, CSRF, OAuth grants, SELinux, SSH host keys and CSP.
• Serverless 11Lambda timeouts, package size limits, VPC networking and cold starts.
• Shell 13Command not found, permissions, quoting, expansion and Makefile syntax.
• Storage 13S3 permissions, NFS mounts, quotas, signed URLs and volume attachment.
• Svelte 10Store subscriptions, load functions and server/client boundaries.
• System 26Disk space, systemd units, file descriptors, OOM killer and scheduled jobs.
• Terraform 18State locks, provider auth, drift, dependency cycles and plan-time unknowns.
• Testing 18Jest, pytest, JUnit, Cypress and Playwright: fixtures, snapshots and timeouts.
• TLS 24Untrusted authorities, expiry, hostname mismatch, chains and cipher negotiation.
• TypeScript 19Assignability, missing declarations, strict null checks and generic constraints.
• Virtualization 8VirtualBox, VMware, Hyper-V, WSL, KVM and hypervisor conflicts.
• Web3 7Gas estimation, nonce management and reverted transactions.
• WebAssembly 7Compile errors, memory bounds and host binding mismatches.
• WebServer 10nginx, Apache, IIS and Caddy: binding, permissions, rewrites and TLS.
• Windows 10Installer failures, missing runtimes, update errors and permission problems.