Hello. If I want to have one or two systems on my LAN have higher priority during times of congestion and I make a set of upload/download queues for "high priority" with let's say weight = 60 and a set for "low priority" with remaining weight of 40 would that mean that the clients in the high priority queue would get at least 60% of the bandwidth (shared fairly among them, I assume) and the clients in the low priority queue would get at least 40%? Would this be a good way to favor my work from home computer and VOIP device over my kids video streaming? Thanks!
Quote from: omegahelix on November 30, 2025, 10:43:19 PMwould that mean that the clients in the high priority queue would get at least 60% of the bandwidth (shared fairly among them, I assume) and the clients in the low priority queue would get at least 40%?
If you use a weight based based scheduler, the weights work like a ratio. If your BW is 100Mbit the Queue with Wight 60 will get ratio 60% of that BW and a Queue with Wight 40 will get ration of 40% of the BW. If you would like to share equally the BW within the Queue for all devices that are classified into the Queue you need to as well configure MASKs on the Queue.
Keep in mind, even the devices to whom you dont want to give any BW, needs to be schaped. They need to fall into so called default class. Cause if they are not in any of those queues/pipes, they will eat into the shaped ones which will cause you exhaustion of the BW and negatively impact shaped ones.
Quote from: omegahelix on November 30, 2025, 10:43:19 PMWould this be a good way to favor my work from home computer and VOIP device over my kids video streaming? Thanks!
Depends on your usage/use-case. It may be enough from BW related terms but, weight based based schedulers do not handle bufferbloat well. They only manage BW based on weights.
Regards,
S.