List Back Order Strategy (SBO OFF): onPush vs Explicit Reevaluation
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report
Hi everyone,
I’m validating List behavior in FlexSim with SBO unchecked (entries-first: the list iterates entries and, for each entry, prioritizes compatible back orders via the Back Order Queue Strategy). I expected identical assignments regardless of how reevaluation is triggered, but I consistently see different outcomes when it’s triggered by onPush vs by an explicit reevaluation (code or List trigger wity Time Interval).
This model/question is inspired by the solution posted here:
https://forums.autodesk.com/t5/flexsim-forum/list-back-order-strategy-option/m-p/13596928#M91487
Different results with onPush vs explicit reevaluation when SBO is OFF (entries-first)
Setup (minimal):
SBO: OFF (entries-first)
Allow Multiple (t+ε): ON • Reevaluate All Values On Pushes: ON
Pullers: 3 processors with Prio = 3, 2, 1 (Request=Require=1)
Entries: 3 operators with Skill = 3, 2, 1
Push pattern: the current model uses staggered pushes (different creation times) and I also tested fully simultaneous entries; results are the same.
Two reevaluation configurations:
- onPush configuration: at the reevaluation time I perform a dummy push into the list.
- Explicit-reevaluation configuration: reevaluation is triggered without a push, either by a PF calling
List("List1").reevaluateBackOrders() or by the List’s Back Order Reevaluation Events → Time Interval.
Observed:
onPush: P3 → Op3, P2 → Op2, P1 → Op1 (desired).
Explicit reevaluation: P3 → Op1, P2 → Op2, P1 → Op3. (expected)
Questions
With SBO OFF (entries-first), why does the onPush configuration produce a different assignment than the explicit-reevaluation configuration, given the same entries and pullers?
Is there a documented difference in the internal entry-iteration order between the onPush loop and the reevaluateBackOrders()/Time Interval loop? If so, how can we inspect or control that iteration order? (The Entries grid order doesn’t seem to match what the algorithm uses.)
I’ll attach the simplified model and screenshots for both configurations. Thanks!
Version 24.1