Enhanced packet redirect capabilities are disclosed herein for draining traffic to a server. In an implementation, a server in an infrastructure service receives a packet from a stateless load balancer. The packet may comprise a request for content. A user space program on the server determines whether a connection identified in the packet belongs to the server. If the connection belongs to the server, the user space program handles the request for the content. If not, the server forwards the packet to a secondary server in the infrastructure service. The secondary server, to which the connection may belong, can then handle the request.

Techniques for oscillatory complementary network property calibration of a network connection can be implemented by measuring a first network property (e.g., latency or bandwidth output) as a function of bandwidth input and performing statistical analysis to determine a correlation. If a non-zero correlation coefficient is detected, a second network property complementary to the first network property can be measured to determine a first value of the second network property. Likewise, the second network property can be measured as a function of bandwidth input to determine a second correlation which, if positive, may indicate how to determine a second value of the first network property. The first value and the second value can be utilized to determine a third value of a third network property (e.g., network latency and network capacity utilized to determine bandwidth-delay product).

A timer is associated with a packet of a flow from a VM at an application executing in conjunction with a vSwitch in a host system, using a processor assigned to the vSwitch in the host system. At the application, using a counter, a number of packets of the flow that are received and acknowledged in response packets is counted, the response packets being received from a receiver of the flow. At the application, using a period measured by the timer and the number of received packets acknowledged as counted by the counter, a CWND value is computed. The CWND value is applied to the flow at the vSwitch such that the vSwitch transmits, from the flow to a network, only a number of packets up to the CWND value.

Method and apparatus for carrying out the method receiving packets, each of the packets comprising a header and a payload. For a particular packet among the packets, the method includes processing at least the header of the particular packet to determine a flow associated with the particular packet; attempting to determine a payload structure based on the flow, the payload structure associated with transport of coded video data in the payload of the particular packet; and if the attempting is successful, repackaging coded video data contained in the payload of the particular packet into a new packet and forwarding the new packet to an external system or storing the new packet in memory.

A communication device receives, using a connectionless protocol, data transmitted by a transmission terminal. The communication device determines, in accordance with a receivable size of the communication device, a data transmission request in order to receive the data from the transmission terminal.

An apparatus and a method for managing network flow congestion are provided. The method for managing network flow congestion includes: receiving a plurality of packets; identifying whether each packet belongs to a predetermined protocol; distinguishing a plurality of connections of the packets belonged to the predetermined protocol; monitoring a buffer usage of each different connections; and performing a congestion check procedure on the currently monitored connection. In the congestion check procedure, when the buffer usage of the currently monitored connection exceeds a threshold, a congestion processing procedure is triggered.

Enhanced packet redirect capabilities are disclosed herein for draining traffic to a server. In an implementation, a server in an infrastructure service receives a packet from a stateless load balancer. The packet may comprise a request for content. A user space program on the server determines whether a connection identified in the packet belongs to the server. If the connection belongs to the server, the user space program handles the request for the content. If not, the server forwards the packet to a secondary server in the infrastructure service. The secondary server, to which the connection may belong, can then handle the request.

Described is a method for estimating throughput between first and second communication devices, the method comprising: determining maximum bottleneck throughput of a communication link between the first communication device and a third communication device, wherein the communication link between the first and third communication devices applies a common access network as between a communication link between the first and second communication devices; determining Round Trip Time (RTT) between the first and second communication devices; transmitting packet by applying User Datagram Protocol (UDP) from the third communication device to the first communication device; measuring packet loss rate associated with the transmitted packet by monitoring sequence number of the packet; and translating measured packet loss rate to Transmission Control Protocol (TCP) throughput according to maximum bottleneck throughput and RTT.

A system transports a plurality of UDP datagrams from a sending application to a receiving application by creating a TCP tunnel between a TCP sending-end and a TCP receiving-end, encapsulating the datagrams in TCP packets at the TCP transmitting-end, transmitting the TCP packets via the TCP tunnel to the TCP receiving-end over a network using a TCP/IP protocol, and extracting the datagrams from the TCP packet and forwarding the extracted datagrams to the receiving application. The TCP tunnel may provide the same delay and responsiveness as UDP protocol. The TCP receiving-end may detect when a packet is missing and request retransmission when a packet is missing, so that the TCP sending-end retransmits the missing packets. The transmitting of TCP packets to the TCP receiving-end continues when the receiving-end detects a missing packet, so that there is no lag in the forwarding of the extracted datagrams. Retransmitted packets may be discarded.

Methods, systems, and computer-readable media for providing reliable switching between different transport protocols (e.g., TCP and UDP) are presented herein. In some embodiments, a computing platform may stop transmission over a transport protocol. Subsequently, the computing platform may synchronize buffers between a client and a host. For example, the buffer may include acknowledgments for undelivered reliable packets between the client and the host. Next, the computing platform may resume transmission over a different transport protocol. In some embodiments, when establishing an initial connection between a client and a host, connections may be attempted in parallel over multiple transport protocols. After a connection over a first transport protocol is established, if a connection is later established over a second transport protocol, the connection may be transitioned to the second transport protocol. Also, ticketing semantics may be preserved in the switch between the different transport protocols.