Thank you aleksisimell.

But it seem to be different issue. I have 20 stages in my jenkins pipeline and I see only first stage data is sent to influxdb and the above function is getting exited after one loop. Any suggestions? Sorry for my earlier interpretation.

Based on your paste, you're collecting data to `myDataField1` and `myCustomDataFields1`, but never writing those to InfluxDB.

Is the job completing successfully? If it is, then the only thing I can think of right away is that `jsonObject.stages` does not contain 20 objects, but only 2. But I have no experience from JsonSlurper.

aleksisimell Thank you for the suggestion. Job is getting completed successfully and also I seee 20 objects in the json and jsonObject.stages.size() returns 20. Is it related to @NonCPS annotation? If I comment that it iterates over all stage but I get java.io.serilizable exception?

In your global configuration, have you checked the "Job scheduled time as timestamp" checkbox? If so, then you will have only 1 timestamp, which causes you to overwrite your data every time you write to InfluxDB. Unchecking this and adding a short wait will force a new timestamp for each write you do.

aleksisimell ，According to sridattasp, in the "jenkins_data" table, each build has multiple duplicate data? Is there any way to generate a single record in the "jenkins_data" table each time, and multiple records in the "jenkins_custom_data" table?

zyun823 That is currently not possible. You're only able to add keys and values to a single "jenkins_custom_data" measurement. I can check how big the workload would be to change the functionality so that you would be able to send multiple measurements for the same custom data in a single call from InfluxDbPublisher.

aleksisimell ，Thank you for your reply and look forward to this feature. The customDataMap and customDataMapTags support list(map) will be great.

For example:

stageDataMapTagsList = []

stageDataMapTags = [:]

stageDataMapTagsList.add(stageDataMapTags)

customDataMapTags['stage'] = stageDataMapTagsList

How is batch processing represented using the Jenkins influxDBPublisher?

In your GitHub repo I saw that there's an array of JUnit test results being processed: https://github.com/jenkinsci/influxdb-plugin/blob/8244a8d0a703f66778a1c8a64ed26d0622f334dd/src/main/java/jenkinsci/plugins/influxdb/generators/JUnitPointGenerator.java  Can I use some similar API that accepts custom data per test?

return points.toArray(new Point[0]);

private void writeToInflux(Target target, InfluxDBClient influxDB, List<Point> pointsToWrite) {
        /*
         * build batchpoints for a single write.
         */
        try {
            influxDB.getWriteApiBlocking().writePoints(pointsToWrite);

For one thing, is this API exposed for Jenkins pipeline? How can I use that in a Jenkinsfile. Also, Is each of these points stored in InfluxDB as its own row?

With thousands of tests it's not clear what overload of InfluxDBPublisher() I should use for writing multiple rows in the DB for a single data map.

My current use case is something like

testData.each {data ->
  influxDbPublisher( selectedTarget: 'jenkins_data', customData: data)
}

However with hundreds -> thousands of tests this takes forever (literally 20+ minutes). I'd like an equivalent invocation for  batch publishing as mentioned in the documentation for influxDB: https://www.influxdata.com/glossary/batch-processing-explained/

Hi zargold,

I was about to write instructions how to do this with CustomDataMap, but looking at the code and your suggestion, I say this is not supported for custom data. For example Robot Framework results process sub points individually and a similar functionality could indeed be added to CustomDataMap data generation.

Hi aleksisimell,

So to be clear if I use JUnit: I would indicate a junit.xml file to process.
It would process my JUnit output and it convert it to Points that look something like:

[{
test_name: "convertMoney: it converts numbers to dollars",
test_duration: 0.23,
test_status: "pass"
}, {
test_name: "convertMoney: it converts dollars to numbers",
test_duration: 0.54,
test_status: "pass"
}...5000 unit tests]

When storing the JUnit data it will:

• store all 5000 unit tests ("points") in a single/fast API request to InfluxDB
• Each Unit test ("point") will be stored in it's own row in influxDB.

So that I can query something like:

SELECT TOP 3 * FROM "jenkins_custom_data" WHERE test_duration <> nil ORDER BY test_duration DESC;

to get the 3 slowest tests?

But there's no way to provide custom "Points" and batch send those to be stored each as it's own row right?