Log Playground: online log analyzer to compare log files
Open up to 3 log files at once, read them side by side or merged into one timeline, and follow one request through all of them. Files never leave your device.
Open 1 to 3 log files
Pick up to 3 at once: in the file window, hold Ctrl (⌘ on a Mac) or Shift while clicking the files. Or drop them anywhere on this page. Application logs (log4j, logback, Python logging), nginx and Apache access logs, PostgreSQL, syslog and anything else that starts each line with a time all work together, and .gz files open directly.
No files to hand? Try a sample: three services that logged the same failed payment.
Reading happens in your browser: the files are not uploaded and nothing is saved. Up to 3 files of up to 500 MB each; how large works well depends on your device's memory.
About the Log Playground
An online log analyzer that keeps your logs private. Open a log in your browser and it finds timestamps and levels, counts errors and warnings, groups repeated errors into patterns, and lets you search, filter, bookmark and export. Nothing is uploaded: the analysis runs on your device.
When something goes wrong, the story is usually split across several logs: the web server saw a 504, the application logged a stack trace, and the database waited on a lock. Most log viewers open one file at a time. This playground opens several together, puts them next to each other and lines them up by time, so you can read one incident across all of them, or merge them into a single timeline ordered by timestamp.
It suits anyone who has to compare log files from different servers or services: microservices that each write their own log, a load balancer and the app behind it, a primary and a replica database, or the same service before and after a deployment.
Following one problem through several logs
| Tool | What it does |
|---|---|
| Click a line | Every other file scrolls to the line closest in time, marks it, and shows the difference, for example +0.42 s. |
| Link scrolling | Scrolling one file scrolls the others to the same moment. |
| Trace across files | Type a request id, order number or user, or select it in a line, and every file shows only the events that contain it. |
| Merged by time | All files in one list, in time order, each line marked with its file's colour. |
| Timeline | One lane per file with errors in red. Drag across it to keep only a time window in every file. |
| Clock offset | When a server's clock runs ahead or behind, press ⚙ Options on that file and shift its times so the files line up. |
Finding what matters
| Tool | What it does |
|---|---|
| Search words | Every word must appear in an event, in any order. "quotes" keep a phrase together and -word hides events that contain it. Aa matches case and .* takes a regular expression. |
| ≋ Patterns | Groups lines that differ only in numbers, ids, addresses and times, with a count and when each first and last happened. Show one pattern, or hide the noisy ones. |
| Bookmarks | Click a line number to bookmark it. ★ Bookmarks lists them from every file in time order; click one to go back to it. |
| ⇩ Export | Save your bookmarks or the lines shown now as text, CSV or JSON, from one file or all files merged by time. |
| Marks by the scroll bar | A thin strip beside each list shows where matches (yellow), errors (red) and bookmarks (blue) are. Click it to jump there. |
| Skip lines | Leave out a start-up banner or header without times, under ⚙ Options. The file points it out when it finds one. |
Making long logs easier to read
| Setting | What it does |
|---|---|
| Wrap | Long lines continue on the next rows instead of running off to the right. Set it for each file. |
| Numbers | File line is the line number in the file. Event numbers each log entry once, so a 40-line stack trace counts as one. Row counts what is shown now, after filters. |
| Colours | Click a file's coloured dot to change its colour, and to highlight lines that contain words you choose, such as timeout in orange. |
| Levels | ERROR, WARN and the other levels are found and coloured, and each can be hidden. Counts are of events, so a stack trace is counted once. |
Questions
Are my log files uploaded?
No. The files are read by your browser, on your device, and are not sent anywhere or saved. Closing the tab clears them.
Which log formats work?
Plain-text logs whose lines start with a time: Java (log4j, logback), Python logging, nginx and Apache access logs, PostgreSQL, syslog, and most others. Times written as ISO 8601, 12/Oct/2026:09:10:00, Oct 12 09:10:00, 2026/10/12, 12/10/2026 or a bare 09:10:00 are found automatically. Rotated logs packed as .gz open directly. JSON logs are not supported yet.
How do I find the most common errors in a log?
Press ≋ Patterns on a file. Lines that differ only in numbers, ids, IP addresses and times are counted together, so 66 "slow response from payment provider: # ms" warnings show as one row with its count and when it first and last happened. Show lists only that pattern; Hide removes it, which is a quick way to clear out noise.
Can I search for several words, hide lines, or use a regular expression?
Yes. Every word you type must appear in an event, in any order; put a phrase in "quotes" to find it as written; start a word with a minus (-healthcheck) to hide events that contain it. Aa matches upper and lower case exactly, and .* treats the whole text as a regular expression. Both the filter box of each file and Trace across files work this way.
How do I save the lines I found?
Click a line number to bookmark it; ★ Bookmarks lists them from every file in time order. ⇩ Export saves either your bookmarks or the lines shown now as plain text, CSV or JSON, from one file or all of them merged by time.
My log starts with lines that have no time. What can I do?
When the first lines of a file have no time, the file says so and offers Skip them. You can also set how many lines to skip under ⚙ Options. Skipped lines are left out of filters, counts, patterns and the timeline.
Is this an online log analyzer?
Yes. It is a free online log analyzer that runs in your browser: open a log file and it finds the timestamps and levels, counts errors and warnings, groups repeated errors into patterns, and lets you filter, search with regular expressions, bookmark and export lines. Unlike most online log analyzers, your file is not uploaded to a server; it is read on your device, so production logs with customer data stay with you.
Can I compare several log files at once?
Yes. Open up to 3 files, for example the web server, the application and the database log, and they appear side by side. Click a line in one and the others jump to the same moment, or choose Merged by time to read all of them as one timeline with each line marked by its file. Most log viewers open one file at a time; this one is built for reading several together.
How do I merge log files by timestamp?
Open the files and choose Merged by time. Lines from every file are put in one list in time order, and filters, the trace and the time window still apply. When servers write times in different zones, or one clock runs a few seconds off, set the zone or a clock offset for that file under ⚙ Options so the merged order is right. ⇩ Export saves the merged lines as text, CSV or JSON.
How do I correlate logs from several services?
Find a value they all log, such as a request id, trace id or order number, and type it into Trace across files: each file then shows only the events that contain it. When a log does not carry the id, as with many database logs, click the failing line in another file and that log scrolls to the same moment.
How do I compare two log files by time?
Open both, then click a line in one of them. The other file scrolls to the line closest in time and shows the difference. Turn on Link scrolling to keep them together as you scroll, or choose Merged by time to read them as one list.
Does it show how many times a search was found?
Yes. Under each file you see how many times the text occurs in the lines shown and in how many events, for example 37 matches in 12 events. Trace across files also shows the total and the count for each file. Use the ▲ ▼ arrows beside the filter box, or Enter and Shift+Enter in it, to jump from one match to the next.
How do I follow one request through several logs?
Type its request id, order number or any other value into Trace across files, or select it in a line and press Trace in all files. Every file then shows only the events that contain it, with stack traces kept attached.
The times in two files are a few seconds apart. Why?
Usually one server's clock runs ahead or behind, or one log is written in another time zone. Press ⚙ Options on that file and set a Clock offset (for example -2.5s) or the zone its times are written in.
What do Event numbers mean?
Event numbering gives each log entry one number, so a stack trace that spans 40 lines still counts as one event. File line numbers count every line, and Row numbers count only what is shown after filtering.
Can I open several log files at once?
Yes, up to 3. Press Open log files and, in the file window, hold Ctrl (⌘ on a Mac) or Shift while clicking the files, or drag them onto the page together. You can also add files one at a time.
How large a log can I open?
Up to 3 files of up to 500 MB each. A one-million-line file opens in a second or two on a typical laptop; very large files need enough free memory on your device.