ADWAN7.1 scheduling issue: under excessive load conditions, high precedence services may also be scheduled away

ADWAN7.1 scheduling issue: on-site operations include both high precedence and low precedence services. Under link overload conditions, even high precedence services may be scheduled away.

ADWAN7.1 scheduling issue: the site has both high precedence services and low precedence services. Under link overload conditions, high precedence services may also be scheduled away. The site operates in a primary-backup path mode.

View the scheduling records. The main reason for scheduling is link overload. For example, the real-time bandwidth of link SJZ-H3C_SR6608M-RT01 To LF-H3C5660-RT02 Link1 IPv6 is 29997 kbps, while its maximum supported bandwidth is 27000 kbps, so this application was scheduled out. The reason is that the current adwan7.1 scheduling principle states that when a link is overloaded, if a high-precedence service experiences a traffic burst, the device will determine that the link cannot accommodate the high-precedence service. To free up space for this service, all other low-precedence services must be scheduled away. However, if only this high-precedence service is scheduled away, the link load will return to normal. Therefore, the high-precedence service occupying significant bandwidth is scheduled out.

The current mechanism in adwan7.1 is normal. On-site observation shows that high precedence traffic bursts last only a few minutes each time. To reduce such scheduling, you can change continuous optimization to cycle optimization.