Tourlane Tech wants to give this talk

Ensure that your command only run once at the same time... WORLDWIDE.

We have some tasks that are trigger every couple of minutes. In case the cluster was slow, we had the problem that new tasks was scheduled before the previous one being done. Making the cluster exponentially slower. So we sit back and think. What can we do to actually prevent this from happening?

Become a patron
Fork me on GitHub!