What a running integration is doing

Camel 3 · 4:04

When a step of the hello integration starts failing, its logs, its failure counts and the trace of one message show what happened, in the terminal and in the admin UI.

Download the mp4Recorded on 2026.2

What it covers

A step that throws on every third message, published as 0.1.2
Checking that it runs, and its failed count, with weikio status
Following and filtering log lines with weikio logs by level and by query
Failed traces and the step that threw, with weikio traces
The Overview, Logs and Metrics sections of the integration in the admin UI
A failed trace as a waterfall, linked to its log lines

Commands in this video

Terminal

$ weikio run ./hello
$ weikio status sandbox/hello
$ weikio logs sandbox/hello --since 10m
$ weikio logs sandbox/hello -f
$ weikio logs sandbox/hello --level error
$ weikio logs sandbox/hello --query 'every third'
$ weikio traces sandbox/hello
$ weikio trace 56ec36911630f4d6da08910969d88f9f

Transcript

0:02Video two left sandbox/hello running, on version 0.1.1.
0:09This video reads what that integration is doing while it runs.
0:16sandbox/hello runs on the agent dev, and logs a line every three seconds.
0:22An integration writes log lines and counts its own messages.
0:27Each message also leaves a trace: the path it took through the route's steps.
0:32The three together are called telemetry.
0:35Telemetry is on for every integration, unless its files turn it off.
0:42To have something to find, this video adds one step to the route.
0:47The step throws an exception when the timer's counter divides by three.
0:55The loop from video two publishes the folder again.
0:58run publishes and deploys in one command, so version 0.1.2 goes straight to devtest.
1:08The following shows both: the route's line,
1:11and the failure every third message hits.
1:14Every third message ends in an error line.
1:21status answers the first question: is it running?
1:24It runs the latest version in devtest, on agent dev.
1:28Under the table, the last hour in one line.
1:32Its failed count is the failing step's work.
1:39logs reads the lines of the last ten minutes.
1:42Each line shows the moment it was written, at its level.
1:46The route and the agent stand between them and the message.
1:50Follow prints new lines as they arrive.
1:53The same error line comes around every nine seconds.
2:01The level filter shows the lines at a level and above.
2:05At error, only the failing lines remain.
2:08The query filter is a full-text search in the log store's own query language.
2:17The traces command lists the messages the integration handled, newest first.
2:22The error column names where a failed trace failed.
2:25One failed trace prints as its steps.
2:29The step that failed carries the exception, and the Failed at line says where.
2:34The note under it gives the command that reads this trace's log lines.
2:43The same reading lives in the admin UI.
2:46The Overview's last hour strip counts the messages and the failures.
2:50The failed tile counts, with the share beside it.
2:54The Logs section draws the same lines as a histogram, with the same filters.
2:59The live tail button follows the same stream.
3:06Metrics draws the messages over time, ok against failed.
3:11Processing time and memory stand beside it, split by route.
3:16The traces section lists the same traces the terminal listed.
3:21A failed trace opens as a waterfall.
3:24It shows where the message failed, and why.
3:27A log line opens its trace, and a trace links back to its log lines.
3:36Here is the whole picture.
3:38The edit became version 0.1.2, and it is what runs.
3:43The JVM's telemetry lands in the hub's three stores.
3:47The terminal and the admin UI read it through the hub.
3:51Video four moves the values out of the files.

Next in this track

Camel 4 · 3:56 Where the values live