Network Management

Database Maintenance fails for CPU Multi Load tables when weight is rounded to 0 in NPM 2026.1

Database maintenance can fail on Orion CPU Multi Load tables when the calculated Weight value is stored as 0 in some records. This is caused by a known issue where, in certain cases, a very small time difference between polls is rounded down to zero. As a result, DB maintenance for the affected tables does not occur until the bad records are corrected or a product fix is applied.

First published date

6/12/2026 8:50 PM

Last published date

6/12/2026 8:50 PM

Overview

Customer running NPM 2026.1 reported database maintenance failures on one Orion instance. Investigation by engineering identified that, in some cases, when the difference between polls is less than one second, the code rounds the value to 0, which leads to Weight = 0 in CPU Multi Load data. 

The issue affects these tables during maintenance review and cleanup activities:

  • CPUMultiLoad_CS_cur

  • CPUMultiLoad_CS_Detail_hist

  • CPUMultiLoad_CS_Hourly_hist

  • CPUMultiLoad_CS_Daily_hist

Product section

Orion Platform

Cause

The root cause is a known issue, where the time difference between polls is less than one second, the code may round that difference to 0. This results in Weight being set to 0 even when PollInterval and RediscoveryInterval are not set to 0.

Resolution

This issue will be addressed in a future release of the SolarWinds Platform. Watch this article for updates.


As a workaround, perform the steps in Error "Divide by zero encountered" During Database Maintenance – CPULoad_CS_Detail_hist / CPUMultiLoad_CS_Detail_hist – SolarWinds Platform.