Check Jenkins job status after triggering a build remotely

jenkins build jenkins-plugins jenkins-cli
Nam Nguyen picture Nam Nguyen · Feb 4, 2015 · Viewed 53k times · Source

I have a script to trigger a job on Jenkins remotely using a token. Here is my script:

JENKINS_URL='http://jenkins.myserver.com/jenkins'
JOB_NAME='job/utilities/job/my_job'
JOB_TOKEN='my_token'

curl "${JENKINS_URL}/${JOB_NAME}/buildWithParameters?token=${JOB_TOKEN}"

After I run it, I get following response:

* Hostname was NOT found in DNS cache
*   Trying 10.5.187.225...
* Connected to jenkins.myserver.com (10.5.187.225) port 80 (#0)
> GET /jenkins/job/utilities/job/my_job/buildWithParameters?token=my_token HTTP/1.1
> User-Agent: curl/7.37.1
> Host: jenkins.myserver.com
> Accept: */*
> 
< HTTP/1.1 201 Created
* Server nginx/1.6.2 is not blacklisted
< Server: nginx/1.6.2
< Date: Tue, 03 Feb 2015 23:40:47 GMT
< Content-Length: 0
< Location: http://jenkins.myserver.com/jenkins/queue/item/91/
< Connection: keep-alive
< Cache-Control: private
< Expires: Wed, 31 Dec 1969 16:00:00 PST
< 
* Connection #0 to host jenkins.myserver.com left intact

I noticed that it returns the queue url in the header: http://jenkins.myserver.com/jenkins/queue/item/91. But I don't know how I should use this return url.

1) I am wondering if anyone knows how I can check the status for the job that I just created?

2) Since the above response does not return the job #, I can't really use this api call: 

curl http://jenkins.myserver.com/jenkins/job/utilities/job/my_job/8/api/json

to check for status. So where can I get the job name and job number after I get the location url from the above response?

Thanks

Answer

P. Hawkins picture P. Hawkins · Jan 30, 2018

I solved this problem using polling of the Jenkins server. When a job is started remotely the return headers has the job queue URL. Using this one can make more API calls to get status.

Steps:

  • start the job
  • parse return 'Location' header
  • poll the queue looking for job to start
    • job queue entry will have an 'executable' entry in its json or xml with job number once it starts
  • poll the job status waiting for a result

I used python and the Requests module to do this

#!/usr/bin/python

import requests
import re
import sys 
import json
import time

# secs for polling Jenkins API
#
QUEUE_POLL_INTERVAL = 2 
JOB_POLL_INTERVAL = 20
OVERALL_TIMEOUT = 3600 # 1 hour

# job specifics: should be passed in
auth_token = 'buildmaster:173223588624f980c3AAA68d4d8efe0c'
jenkins_uri = '192.168.115.187:8080'
job_name = 'rf_systest'
build_token = 'rf_systest_auth_token'

# start the build
#
start_build_url = 'http://{}@{}/job/{}/build?token={}'.format(
        auth_token, jenkins_uri, job_name, build_token)
r = requests.post(start_build_url)

# from return headers get job queue location
#
m = re.match(r"http.+(queue.+)\/", r.headers['Location'])
if not m:
    # To Do: handle error
    print "Job starte request did not have queue location"
    sys.exit(1)

# poll the queue looking for job to start
#
queue_id = m.group(1)
job_info_url = 'http://{}@{}/{}/api/json'.format(auth_token, jenkins_uri, queue_id)
elasped_time = 0 
print '{} Job {} added to queue: {}'.format(time.ctime(), job_name, job_info_url)
while True:
    l = requests.get(job_info_url)
    jqe = l.json()
    task = jqe['task']['name']
    try:
        job_id = jqe['executable']['number']
        break
    except:
        #print "no job ID yet for build: {}".format(task)
        time.sleep(QUEUE_POLL_INTERVAL)
        elasped_time += QUEUE_POLL_INTERVAL

    if (elasped_time % (QUEUE_POLL_INTERVAL * 10)) == 0:
        print "{}: Job {} not started yet from {}".format(time.ctime(), job_name, queue_id)

# poll job status waiting for a result
#
job_url = 'http://{}@{}/job/{}/{}/api/json'.format(auth_token, jenkins_uri, job_name, job_id)
start_epoch = int(time.time())
while True:
    print "{}: Job started URL: {}".format(time.ctime(), job_url)
    j = requests.get(job_url)
    jje = j.json()
    result = jje['result']
    if result == 'SUCCESS':
        # Do success steps
        print "{}: Job: {} Status: {}".format(time.ctime(), job_name, result)
        break
    elif result == 'FAILURE':
        # Do failure steps
        print "{}: Job: {} Status: {}".format(time.ctime(), job_name, result)
        break
    elif result == 'ABORTED':
        # Do aborted steps
        print "{}: Job: {} Status: {}".format(time.ctime(), job_name, result)
        break
    else:
        print "{}: Job: {} Status: {}. Polling again in {} secs".format(
                time.ctime(), job_name, result, JOB_POLL_INTERVAL)

    cur_epoch = int(time.time())
    if (cur_epoch - start_epoch) > OVERALL_TIMEOUT:
        print "{}: No status before timeout of {} secs".format(OVERALL_TIMEOUT)
        sys.exit(1)

    time.sleep(JOB_POLL_INTERVAL)

Output:

Tue Jan 30 16:24:08 2018: Job rf_systest added to queue: http://buildmaster:[email protected]:8080/queue/item/164/api/json
Tue Jan 30 16:24:39 2018: Job rf_systest not started yet from queue/item/164
Tue Jan 30 16:25:00 2018: Job started URL: http://buildmaster:[email protected]:8080/job/rf_systest/79/api/json
Tue Jan 30 16:25:01 2018: Job: rf_systest Status: None. Polling again in 20 secs
Tue Jan 30 16:25:21 2018: Job started URL: http://buildmaster:[email protected]:8080/job/rf_systest/79/api/json
Tue Jan 30 16:25:21 2018: Job: rf_systest Status: None. Polling again in 20 secs
Tue Jan 30 16:25:41 2018: Job started URL: http://buildmaster:[email protected]:8080/job/rf_systest/79/api/json
Tue Jan 30 16:25:41 2018: Job: rf_systest Status: None. Polling again in 20 secs
Tue Jan 30 16:26:01 2018: Job started URL: http://buildmaster:[email protected]:8080/job/rf_systest/79/api/json
Tue Jan 30 16:26:01 2018: Job: rf_systest Status: None. Polling again in 20 secs
Tue Jan 30 16:26:21 2018: Job started URL: http://buildmaster:[email protected]:8080/job/rf_systest/79/api/json
Tue Jan 30 16:26:21 2018: Job: rf_systest Status: SUCCESS

JSON from a Jenkins queue once its job has started:

{
    "_class": "hudson.model.Queue$LeftItem",
    "actions": [
        {
            "_class": "hudson.model.CauseAction",
            "causes": [
                {
                    "_class": "hudson.model.Cause$RemoteCause",
                    "addr": "10.20.30.60",
                    "note": null,
                    "shortDescription": "Started by remote host 10.20.30.60"
                }
            ]
        }
    ],
    "blocked": false,
    "buildable": false,
    "cancelled": false,
    "executable": {
        "_class": "org.jenkinsci.plugins.workflow.job.WorkflowRun",
        "number": 45,
        "url": "http://192.168.115.187:8080/job/rf_systest/45/"
    },
    "id": 95,
    "inQueueSince": 1517342648136,
    "params": "",
    "stuck": false,
    "task": {
        "_class": "org.jenkinsci.plugins.workflow.job.WorkflowJob",
        "color": "blue_anime",
        "name": "rf_systest",
        "url": "http://192.168.115.187:8080/job/rf_systest/"
    },
    "url": "queue/item/95/",
    "why": null
}

Related questions

How to add String parameter to Jenkins with optional checkbox

I'm trying to add a String parameter to my jenkins build, but i can't find an option to make it optional, on the Jenkins WIKI i found a screeshot and there was an option to make it optional: https://wiki.…

Read More
How can I modify the Load Balancing behavior Jenkins uses to control slaves?

We use Jenkins for our CI build system. We also use 'concurrent builds' so that Jenkins will build each change independently. This means we often have 5 or 6 builds of the same job running simultaneously. To accommodate this, we have 4 slaves …

Read More
How schedule build in Jenkins?

How do I schedule a Jenkins build such that it would be able to build only at specific hours every day? For example to start at 4 PM 0 16 1-7 * * I understand that as: 0 minutes, at 4 o'clock PM from Monday to Sunday …

Read More