Running into a cfscheduler issue it seems. We have a job that takes 7 minutes to run, scheduled to start every 10 minutes. In looking at the http.log and cfscheduler.log, everything looks normal:

"Information","DefaultQuartzScheduler_Worker-1","05/06/25","07:40:00","","Starting HTTP request {URL='<https://www.xyz.com/app/jobs/college/generate_daily_files.cfm>', method='get'}"
"Information","DefaultQuartzScheduler_Worker-1","05/06/25","07:47:57","","HTTP request completed  {Status Code=200 ,Time taken=477114 ms}"

"Information","DefaultQuartzScheduler_Worker-1","05/06/25","07:40:00","","Task SDQ.SDQ - generate_daily_files.cfm triggered."
"Information","DefaultQuartzScheduler_Worker-4","05/06/25","07:50:00","","Task SDQ.SDQ - generate_daily_files.cfm triggered."

However, the job actually runs 2 additional times about 2 minutes after the first (correct) kicks off, and then about an additional 2 minutes. We added a simple email to send start time, end time, browser agent and a url flag:

1	0506_082200 Started  (this is the scheduled start of the job)
2	0506_082200 Browser Agent = CFSCHEDULE
3	0506_082200 URL.job = cdf 

1	0506_082357 Started (not scheduled)
2	0506_082357 Browser Agent = CFSCHEDULE
3	0506_082357 URL.job = cdf

1	0506_082657 Started (not scheduled)
2	0506_082657 Browser Agent = CFSCHEDULE
3	0506_082657 URL.job = cdf

These two extra runs are not logged in http.log or the cfscheduler log file. If you run it directly from the URL it runs correctly - one and done as expected. Any ideas on why the task scheduler seems to be running this job extra times and how we can get around that?

does this just start to happen and only every few days, or every day that the job runs?

It's been running amok for a while (several months). We added a url variable with a timestamp in it, so if actually fired multiple times that would change - and it doesn't. The only other oddity we've found is that the main/correct job kicks off at :00 seconds, the subsequent runs that shouldn't happen both always occur at :58 seconds.

it might be worth adding logging every few steps to see if there is something happening at the 1m57s mark. Another question - if you turn off the scheduled job, do the 2 anomalous ones still run? If you tell the job to run manually from the task scheduler, do you get 3 jobs running still?

Just some questions to clarify: • What were the three end-times (in the email)? • How do you know that the two extra processes are additional runs of the job? • Might it be that you are looking at two helper processes generated by the scheduler?