HQoS Implementation on Juniper MX80 / MX104 Fixed Ports

    Widespread news in narrow circles was the appearance in the Junos 16.1 release of extended queuing mechanisms on fixed 10G ports on the MX80 / MX104 platforms.

    As it sometimes happens, technical documentation on new features comes out with some delay, even now the sections devoted to this topic leave a feeling of understatement. Fortunately, colleagues from the Juniper Moscow office at the Juniper Networks Summit helped find answers, thanks a lot to them.

    The resources of the QX Trio chip of the first generation (namely, these platforms are built on it) are not distributed between interfaces on modular cards in a trivial way, which means that the manufacturer’s declared subscriber density per chassis can be obtained only under certain conditions.

    The buffer memory of the QX chip is divided between two blocks of schedulers, each of which is capable of serving 32,000 queues. In turn, each scheduler works with two groups of interfaces on different modular cards, providing each group with no more than half of its queues.

    image

    In the case when there is only one module card in the chassis, each scheduler block works with a separate group of ports on this card in exclusive mode, thereby increasing the available number of queues.

    image

    In the case of fixed ports, the situation is somewhat different - ports are divided into pairs (0, 2) and (1, 3), and each pair of ports works with one of two schedulers. Those. each pair of ports is capable of serving no more than 32,000 queues.

    From here a few simple rules follow to achieve the declared subscriber density per chassis:

    • if you land subscribers on two modular cards - divide them by cards and by port groups on the card;
    • if there is one modular card in the chassis, divide the subscribers into port groups;
    • if you land on fixed ports, divide them into port groups.

    Also popular now: