locked
Web jobs schedule has suddenly stopped on 20 Jan... ?? RRS feed

  • Question

  • I've had a recurring hourly webjob that I created in VS, and it's been working fine for the past year. In the Classic Portal it always showed up as "scheduled > hourly", and it was run successfully.

    But I checked now on Jan 20th 2016 and see that it hasn't run for over a day.

    * When I go to the Classic Portal, it still shows up as "scheduled > hourly", but it hasn't been run for over a day. When I click "Run Once" it says "Could not run <myjob>. Please try again. If the problem persists, please contact support'. The classic portal shows the webjob status as "Starting..." indefinitely, until I re-open the webjob page, whereupon it shows the status as "Enabled".

    * When I go to the new portal, it no longer calls my webjob "scheduled". Instead it calls it "on-demand".

    * In the new portal, I did "Browse > Scheduler Job Collections" and tried to create a new job. I copied down the job's activation technique from the classic portal: it was an https post to https://<mything>.scm.azurewebsites.net/api/triggeredwebjobs/<myjob>/run with two headers...

       header "authorization" value "Basic JHBv...xyz"
       header "content-type" value "text/plain"

    So I set up these two headers in the new portal's scheduler job configuration. But whenever it runs, either by the "run now" button or by its schedule, it produces an error message...
    Response from host '<mything>.scm.azurewebsites.net': 'Conflict' Response Headers: Pragma: no-cache
    ...
    Body: Cannot start a new run since job is already running.

    What gives?

    Q1. Why does it say "job is already running" when the job emphatically isn't running? (it's a fast job, lasts about 10 seconds).

    Q2. Was it the right thing for me to do, to set up a new scheduler in the portal?

    Q3. What is the correct way to set up a job? I just blindly copied the https:// uri, and the authorization header, from the classic portal. But it seems like there must be a better way to learn what exactly is the way to trigger a webjob. And once I learn it, then I should be able to trigger the webjob myself from my own desktop machine by sending the appropriate http request to the appropriate endpoint, right?

    Q4. I wasn't sure about that authorization header. I noticed that the scheduler let me click on "authorization" and chose basic/advanced, and it asked for a username and password. But I don't know what that username and password should be.

    It's a pain that all this has changed under my feet! I understand that Azure wanted to push out its new CRON way of scheduling jobs. But I'm disappointed that it seems to have broken my existing webjobs, and that I have to go back and get them all working again somehow.

    Thanks in advance for any advice!

    Sunday, February 21, 2016 5:44 AM