nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Any problems with jobs that run

Bowker

We receive feeds (via FTP) throughout the day. The workflow waits 5 minutes, checks to see if we have a new transmission (external task) if yes - then process the files and go back to sleep for 5 minutes and check again. If the files are not ready - just sleep for 5 minutes.

Here's the question. Every 5 minutes I have a dummytask and an externaltask that transition back and forth. Will the workflow engine care that this job has thousands of transitions (responsiveness, memory utilization, disk utilization...)?
(near 288 iterations; 596 transitions per day; ~4000 per week; ~100,000 annually)

Find more posts tagged with

Comments

Migrateduser

I heard something about the workflow subsystem killing a job after about 1000 transitions (thinking the job had gone crazy).

Migrateduser

Yes, John is correct. There is a limit on the number of transitions that can occur within one job (to prevent runaway loops), and it is currently hardwired to 1000.

This limit will <i>probably</i> become configurable in a future release, but this still would not seem like an appropriate use for a workflow job. The workflow system needs to maintain the history of all of those trivial transitions. It would cleaner to use something like a cron job to perform your test, and then initiate a workflow job under the appropriate condition.

Brinko Kobrin
Interwoven Staff Engineer