Skip to main content

Hi IFS Community!

Reposting this to the Framework & Experience subforum. Here’s the contents of the original post to the Buying sub-forum:

I’ve been tasked with hunting down why a Scheduled Database Task for process Mrp_Process_API.Schedule_Mrp_Process__ keeps ending with a status of ‘Error’. The Error Text field has a notice “This background job was set to state Error because the executing Oracle process was killed. ( Detected by cleanup process at 2025-01-08-23.49.11 )”. I’ve checked out a few other Community posts noting that the Light Cleanup process may be responsible, but the Light Cleanup job was posted / started executing 2 seconds after the time reported in the MRP background job. The last bit of output in the background job is:

“Site X MS Calculation for MS parts that have received exploded demand started at 20250108 21:36:24”

I’ve checked out the Oracle Alert log and other background jobs executing around the same time that the process was killed, but the only thing that I see being suspicious is this entry in the Oracle Alert log:

2025-01-08T23:54:59.981027-06:00
NET  (PID:13564): Possible network disconnect with primary database
NET  (PID:13564): while processing B-1160408753.T-1.S-56173 BNUM:2138112 BCNT:939
NET  (PID:13564): RFS client TEST (PID:0)

I know I’m taking this in from a technical perspective, but figured out reach out to some functional folks instead of posting this in the Framework & Experience sub-forum to see if anyone else has encountered this and identified a resolution.

Many thanks, and let me know if there’s anything else I could supply to help with identifying a resolution.

Regards,

Randy

Hi ​@EsiRandyG 
It will be good if you could:

Analyze Oracle Alert Logs and Trace Files

  • Check for additional details related to the network disconnect or other errors logged around the same time as the job failure.
  • Look for Resource Manager or session termination entries.

Validate MRP Job Configuration

  • Review the settings for Mrp_Process_API.Schedule_Mrp_Process__:
    • Ensure the data scope is appropriate for the job (e.g., reducing the number of sites or parts processed in a single run).
    • Check for any customizations or modifications to the MRP process that might increase its resource demands.

It was determined to be a data / configuration issue. Here’s the findings from the customer w/ the resolution:

During our year-end process for updating all standard costs for all products at all sites, we use the multi-site planning checkbox to run inter-co costing when a part is sourced from another site as Primary supplier.  There are over 13K primary suppliers for inter-co across all sites and so I manage these updates for the cost accountant during the year end cost process. 

However, if a part is checked with multi-site planning at the time that MRP is run, the MRP will fail as it is looking to create Distribution Orders and in general handles the procurement differently.  Once I realized that the parts were still checked for multi-site planning, I removed them and manually reprocessed all the MRP runs per site, and they completed successfully.  Since I used the Scheduled jobs and just ran them directly, I’m certain we won’t have the issue again this evening but will update everyone tomorrow.

I don’t know why having all the parts checked for multi-site would cause Oracle to shut down the MRP, but it did happen on the first job I ran manually from the scheduled job before I removed multi-site.  The Oracle Error did not appear when I ran it the second time. 

Thanks ​@truth_trump for writing a reply, it started getting us to think in a data-construct mindset instead of a system issue.

Regards,

Randy

Reply


Terms & ConditionsCookie settings