This didn’t work either.
In my opinion, you need to hire a consultant who can help you ensure your resources are properly configured, and your setup files support your goals. There is something wrong with your data (or possibly your system configuration).
This is why I’ve posted screenshots of everything in the system, resource, resource group, and job configuration for scheduling and am working with trying to schedule a single resource. This was originally set up by a consulting firm and has never worked properly. The goal is to figure this out so that we don’t have to hire a consultant every time we want to make a change to anything. I was also hoping that maybe in the future the information in this thread would be of use to someone else who ends up in my situation in the future. I greatly appreciate all your help in this thread and will continue to work towards a solution that doesn’t involve having to pay someone else to come in and make the software we are already paying for work. If and when I solve this, I will post here my findings so others can hopefully have success in the future.
I would call that fellow back and tell him he did not finish his work.
Unfortunately, this was done a few years ago before I was with the company and from what I am told, the person who set all this up no longer works for that firm and the ones there are clueless. Although, looking at the scheduling results, not so sure he wasn’t clueless as well.
I feel for you. You’re really close to where you want to be, there is no good reason for you to be getting the results you are seeing. Good luck with your issue.
1 Like
It seems like you are using existing data for the jobs. Have you tried creating new jobs and scheduling them? In the last screen shots all of the other operations look complete.
Also, just curious if you have already tried unapproving, removing those existing jobs from the schedule and approving/rescheduling? Reason I ask, I’ve had some issues in the past with some jobs preexisting on the schedule, that I could only clear up by manual remove/re-adding.
1 Like
I have not created new jobs. This is old data. I was hoping to get it to work with scheduling existing jobs but next on my list was to close out all current jobs and start new ones to see what happens. Will keep this thread updated with results.
@TDray , just out of curiosity, have you ever opened a case with Epicor Support for this issue? I didn’t see it mentioned in the thread, but perhaps you have and I just missed that point. Also, have you ever considered requesting a “Customer Value Exchange Workshop”? They are (or at least were) free, and I presume that Epicor still offers them. You could specify scheduling as a key issue for review during the workshop.
I have not opened a case with Epicor support due simply to the lack of support that I have read about scheduling but might be worth it.
I have not heard of the Customer Value Exchange Workshop before.
Here’s a link to a separate discussion on the workshops…
I created a new resource group. I moved this resource into it. I then manually went through and changed all the operations that had pointed at this resource to specifically point at another resource. Created new jobs with operations using the new resource group. Generate Shop Capacity, Calculate Global Scheduling Order, Global Scheduling.
Resource Scheduling Board looks great. Jobs scheduled finitely with no overlaps. Perfect. Just what I’ve been looking for.
Query that pulls all records from ResourceTimeUsed table for this resource has two records that do not show up on the resource scheduling board and would appear to be overscheduling. However, the date created on them (SysDate) are old. One from August and one from September. Apparently the table is not being cleaned up very well and old data persists in it. Something to think about when accessing this table. For example, I plan on accessing this table to create reports. Have to put a check on the SysDate column to ensure that only relevant data is collected. ResourceGrpID is somehow showing the new resource group which had not been created when these records were written. Appears this field may be calculated even though it appears to be stored.
Overload Informer
This is where the real confusion lies. According to this, the resource is overscheduled on October 5th. It also shows it in the old resource group. There are no records in the ResourceTimeUsed table for this on the 5th. Even the two incorrect entries in the ResourceTimeUsed table are not for this date. So where does the overload informer actually pull its data from?
To summarize, creating a new resource group and new jobs with operations that pointed at the new resource group seemed to cause scheduling to work for the most part correctly. Exceptions being that old records in the ResourceTimeUsed table are not all cleaned up. This can be worked around by checking the SysDate. The Resource Scheduling Board and the Multi Resource Scheduling Board appear correct. The Overload Informer is a complete mess that appears to pull its data from somewhere completely different than all other scheduling resources and is pulling old data rather than the new data.
I have never gotten meaningful results from Overload Informer.
1 Like
After doing some research on the overload informer, it seems that is the experience of most. I put together a query that pulls from the ResourceTimeUsed table to check for overscheduling. Using this table and rescheduling those jobs via the job entry > scheduling I was able to eliminate my overscheduling (all but 1 instance where two jobs were scheduled). The overload informer still lists a dozen or so, but none of those show up anywhere else (resource scheduling board, queries, etc.) Currently rerunning global scheduling to see if everything stays finitely scheduled, or whether it messes it all up again.
The scheduling method shouldn’t change, if it does, you selected a flag incorrectly when launching.
We use shop load reports to look for overloads.
It shouldn’t, but it did. I left the flag for finite load checked and also checked ignore locks.
Granted, this is on the old jobs rather than new ones. But having to redo all the jobs on live is something I’m trying to avoid.
Currently it’s only two resources that are still overscheduling so hopefully after digging deeper into those two I can discover something. Not sure why if I schedule directly from the job it moves it out further and schedules finitely but if I use global scheduling, it does what it wants on some jobs.
Are there settings that affect the way global rescheduler works (besides those chosen when you launch it)?
What is your finite horizon on the resources that schedule infinitely?
As far as I know, everything that affects the global scheduling is set correctly. (This thread has screenshots of everything that I believe affects it and shows my settings for these). The finite horizon on those resources are set to 180.