Logs error matcher

What it does

It reads a log line by line and recognizes the errors that stop or break self-hosted services most often: Docker (a port already allocated, the daemon or its socket, an image for another CPU, a missing command, an image that cannot be pulled, no network ranges left), PostgreSQL and MariaDB / MySQL (logins, roles, connections, data from another version, two servers on one data directory), nginx and Caddy (upstreams refused, timed out or not found, configuration errors), certificates (ACME validation, rate limits, expired or untrusted certificates), Node.js and Python (missing modules, the heap limit, Gunicorn timeouts) and Linux itself (the OOM killer, a full disk, a port in use, refused connections, names that do not resolve, read-only file systems, open file limits, permissions). Each error comes with what it means, what to do first, the official documentation, and the wizard or tool of this site that continues from it.

How it works

1. Paste the log or choose a file, up to 8 MiB. It is read on this page in a background worker, so the page stays responsive on a long log; nothing is uploaded, stored or sent to analytics.
2. The log is masked first: IP addresses (except 0.0.0.0, loopback and netmasks), e-mail addresses, tokens and keys, passwords in URLs and values of variables whose names look like secrets. Everything shown, copied, downloaded or handed off comes from the masked log.
3. Each line is compared with a fixed list of known errors, written as patterns that cannot run away on a long or hostile line. The first, most specific match explains the line: an nginx line that says a port is in use is explained as a port in use.
4. The result groups the lines by error, with their line numbers, so you can find them in the original. The same log always gives the same result.

Example

The log of a small Compose stack: the database rejects the app's password, the app tries localhost instead of the database service, then runs out of heap, and nginx answers 502.

db-1     | 2026-10-12 08:14:03.512 UTC [1] LOG:  database system is ready to accept connections
db-1     | 2026-10-12 08:14:09.877 UTC [68] FATAL:  password authentication failed for user "app"
app-1    | Error: connect ECONNREFUSED 127.0.0.1:5432
app-1    |     at TCPConnectWrap.afterConnect [as oncomplete] (node:net:1611:16)
app-1    | notify: sending alert to [email protected]
app-1    | FATAL ERROR: Reached heap limit Allocation failed - JavaScript heap out of memory
proxy-1  | 2026/10/12 08:15:01 [error] 29#29: *1 connect() failed (111: Connection refused) while connecting to upstream, client: 203.0.113.7, server: example.com, request: "GET / HTTP/1.1", upstream: "http://172.18.0.4:3000/", host: "example.com"

What the matcher finds (masked):

Limits

• It recognizes known messages, not causes: one line can be the symptom of an error that a different program logged earlier. Read the lines around each match.
• Only the errors in the list are recognized; a message reworded by a new version may be missed until the list is updated. It is not a log analysis or monitoring system.
• Masking finds common formats of addresses and secrets and can miss others (a host name, a customer id, a secret in an unusual format). Read the masked log before you share it.
• Nothing is sent to an AI model or a server: the explanations are fixed texts, the same for everyone.

Sources

1. Docker Docs: Port publishing and mapping
2. ss(8), Linux manual page
3. Docker Docs: Linux post-installation steps (manage Docker as a non-root user)
4. Docker Docs: Troubleshooting the Docker daemon
5. Docker Docs: Multi-platform builds
6. Dockerfile reference: CMD
7. Dockerfile reference: ENTRYPOINT
8. Docker Docs: docker image pull
9. Docker Docs: docker login
10. Docker Docs: dockerd, daemon configuration file
11. Docker Docs: Prune unused Docker objects
12. PostgreSQL Docs: Authentication problems
13. PostgreSQL Docs: The pg_hba.conf file
14. Docker Official Image: postgres
15. PostgreSQL Docs: Database roles
16. PostgreSQL Docs: Upgrading a PostgreSQL cluster
17. PostgreSQL Docs: Connections and authentication (max_connections)
18. PostgreSQL Docs: Starting the database server (client connection problems)
19. Docker Docs: Control startup and shutdown order in Compose
20. MySQL 8.4 Reference Manual: Troubleshooting problems connecting to MySQL
21. Docker Official Image: mariadb
22. MySQL 8.4 Reference Manual: Can't connect to [local] MySQL server
23. MySQL 8.4 Reference Manual: Running multiple MySQL instances on one machine
24. nginx: ngx_http_proxy_module
25. Docker Docs: Networking in Compose
26. nginx: ngx_http_upstream_module
27. Caddy Docs: Command line (caddy validate, caddy adapt)
28. Caddy Docs: The Caddyfile
29. Let's Encrypt: Rate limits
30. Let's Encrypt: Challenge types
31. Let's Encrypt: Certificate Authority Authorization (CAA)
32. Caddy Docs: Automatic HTTPS
33. Certbot User Guide
34. OpenSSL: openssl-s_client
35. update-ca-certificates(8), Debian manual page
36. Node.js: NODE_EXTRA_CA_CERTS
37. Node.js: Modules, loading from node_modules folders
38. npm Docs: npm ci
39. Node.js: --max-old-space-size
40. Docker Docs: Resource constraints, memory
41. Python Docs: ModuleNotFoundError
42. pip User Guide: Requirements files
43. Gunicorn Docs: Settings, timeout
44. Linux kernel: Control Group v2, memory
45. df(1), Linux manual page
46. errno(3), Linux manual page
47. Node.js: Common system errors
48. Node.js: DNS, implementation considerations
49. resolv.conf(5), Linux manual page
50. mount(8), Linux manual page
51. Docker Docs: Bind mounts
52. getrlimit(2), Linux manual page
53. Docker Docs: docker container run, --ulimit
54. nginx: Command-line parameters (nginx -t)

Related

• Container keeps restarting: find the cause from its exit code
• 502 Bad Gateway behind a reverse proxy: find where the path to your container breaks
• Docker server disk full: find what fills it and free it safely
• Docker volume permissions

Next step

Container keeps restarting: find the cause from its exit code