Docker Logs Grep Guide
Filter docker compose logs with grep-style context, invert and dedupe modes.
A log is a list of messages from a program. A busy container may print too many messages to read one by one. Docker Logs Grep lets you find the lines you care about and the lines next to them. You paste logs into the page; the tool does not fetch logs from Docker for you.
Find a failure in four lines
Paste this made-up example into Paste docker compose logs output:
api-1 | starting
api-1 | connected to database
api-1 | ERROR request failed
api-1 | retrying request
Type error into Search pattern. Leave Ignore case on, so lowercase error also finds uppercase ERROR. Pick Matching lines only under Grep type. The result should contain only the third line. With Line numbers on, it starts with 3:; that means the match was on line 3 of the pasted text, not that Docker assigned it an ID.
Now choose Match with context. Set Lines before to 1 and Lines after to 1. You should see the database message, the error, and the retry. The lines around a match help you see what happened just before and after it. A - after a line number marks context; a : marks the matching line.
If you get no matches
Check the spelling and whether Ignore case is on. Turn Regex off if you want special characters such as . to be matched literally. Regex means a search pattern with special rules; for example, error|timeout finds either word when Regex is on. You can switch to Count matches if you only need a number.
Real logs can contain tokens, cookies, and personal data. Check and redact them before sharing an excerpt with anyone.
Where your input goes
Processing runs in your browser. The tool does not upload your input to ilham.dev.
Related guide
For more background, read Filtering Docker Logs Effectively.
Questions you might have
- How do I get 10 lines before and 5 lines after a match?
- Leave the grep type on Match with context, set Lines before to 10 and Lines after to 5. That is the same as grep -B 10 -A 5, and grep -C would use the larger of the two.
- What is the difference between Regex and plain search?
- With Regex checked, the pattern is a JavaScript regular expression, so error|timeout matches either word and ^api matches the start of a line. With Regex unchecked every character is literal, so a dot matches only a dot.
- Can I dedupe a noisy log?
- Yes. Use the Unique matching lines type, which behaves like grep … | sort | uniq -c and prefixes each distinct line with how many times it appeared. A pattern of . matches every line, so it can dedupe the whole file.