See this 2017 users list post and its attachments; the determination was: "When the standby transition completes, vm2 has more remaining utilization capacity than vm1, so the cluster wants to run sv-fencer there. That should be taken into account in the same transition, but it isn't, so a second transition is needed to make it happen. ... A workaround is to assign some stickiness or utilization to sv-fencer."

Most likely, the problem can occur only when placement-strategy is balanced and some resource has no utilization value.