Abstract: A device may include one or more processors. The device may receive an instruction identifying a set of objects to be generated by a kernel associated with the device. The kernel may generate the set of objects based on receiving information identifying a corresponding set of write operations. The device may provide a first message to cause the kernel to perform first operations corresponding to a first subset of objects of the set of objects. The device may receive one or more notifications indicating whether each operation, of the first operations, was successfully performed. The device may determine, based on whether each operation was successfully performed, a quantity of objects to include in a second subset of objects, of the set of objects. The device may provide a second message to cause the kernel to perform second operations corresponding to the second subset of objects.

Abstract: In general, techniques are described that facilitate scalable wholesale layer two (L2) connectivity between customers and service providers and a demarcation between the L2 wholesale network and one or more ISPs with which customers communicate L2 PDUs. In one example, a network device receives PDU having both a service identifier identifying a service virtual local area network (SVLAN) and a customer identifier identifying a customer VLAN (CVLAN). A virtual switch determines whether an entry of a L2 learning table is associated with both the service identifier and the customer identifier of the PDU. When no such entry exists, a VLAN learning module updates the L2 learning table to create a new entry that maps to a network device interface and is associated with both the service identifier of the PDU and a plurality of customer identifiers that includes the customer identifier of the PDU.

Abstract: Techniques are described that facilitate performing accounting and billing separately for each customer of a service provider account registered to an owner of a shared customer device. An access router performs separate data usage accounting for each of the customers in sub-accounts of the registered service provider account based on customer virtual local area network (CVLAN) and service VLAN (SVLAN) tags included in data packets. A network operator may, therefore, generate a bill for the shared customer device that includes separate data usage charges for each customer. To facilitate the techniques, the access router may determine a CVLAN and SVLAN associated with a customer. The CVLAN and SVLAN tags are included in a service request toward a service provider, and copied into a service request reply toward the shared customer device. The shared customer device then includes the tags in data packets to identify the customer to the access router.

Abstract: In general, techniques are described that facilitate scalable wholesale layer two (L2) connectivity between customers and service providers and a demarcation between the L2 wholesale network and one or more ISPs with which customers communicate L2 PDUs. In one example, a network device receives PDU having both a service identifier identifying a service virtual local area network (SVLAN) and a customer identifier identifying a customer VLAN (CVLAN). A virtual switch determines whether an entry of a L2 learning table is associated with both the service identifier and the customer identifier of the PDU. When no such entry exists, a VLAN learning module updates the L2 learning table to create a new entry that maps to a network device interface and is associated with both the service identifier of the PDU and a plurality of customer identifiers that includes the customer identifier of the PDU.