Dynamic management of packet loss
Exemplary methods, apparatuses, and systems include duplicating a packet within a plurality of packets to be transmitted to a destination computing node as a sequence of packets. The plurality of packets including the duplicate of the packet are transmitted to the destination computing node. Upon receiving a first acknowledgement of the packet from the destination computing node, it is determined that the first acknowledgment is directed to a duplicated packet. In response to determining that the first acknowledgment is directed to a duplicated packet, it is determined that a second acknowledgement has yet to be received for each of one or more packets within the plurality of packets transmitted prior to the packet. In response to determining that the second acknowledgement has yet to be received, the one or more packets are retransmitted to the destination computing node.
1. A computer-implemented method, comprising:
determining an amount of packet loss in transmitting packets to a destination computing node;
determining, based on the amount of packet loss, a duplication pattern for a plurality of packets to be transmitted to the destination computing node as a sequence of packets, wherein the duplication pattern includes (i) a number of candidate packets of the sequence of packets to be duplicated, (ii) a number of duplicates of each of the candidate packets, and (iii) a non-zero spacing between duplicates of each of the candidate packets;
duplicating a first packet within the plurality of packets according to the duplication pattern, including generating one or more duplicates of the first packet and spacing the one or more duplicates within the plurality of packets according to the non-zero spacing;
transmitting the plurality of packets including the one or more duplicates of the first packet to the destination computing node;
receiving a first acknowledgement from the destination computing node; and
in accordance with a determination that the first acknowledgement is directed to a duplicate of the first packet:
in accordance with a determination that a second acknowledgement has yet to be received for each of a first one or more packets within the plurality of packets transmitted prior to the duplicate of the first packet, retransmitting the first one or more packets to the destination computing node.
2. The computer-implemented method of claim 1 , further comprising calculating the number of candidate packets as a predetermined percentage of the amount of packet loss.
3. The computer-implemented method of claim 2 , further comprising, in accordance with a determination that the amount of packet loss is more than a predetermined threshold, capping the number of candidate packets at a predetermined maximum value.
4. The computer-implemented method of claim 1 , further comprising determining the non-zero spacing between the duplicates based on a predetermined minimum number of non-duplicated packets to be transmitted between duplicated packets.
5. The computer-implemented method of claim 1 , further comprising duplicating the first packet after the sequence of packets.
6. The computer-implemented method of claim 1 , wherein the duplication pattern further includes a respective number of the duplicates of each of the candidate packets.
7. The computer-implemented method of claim 1 , further comprising:
subsequent to retransmitting the first one or more packets to the destination computing node, determining a second amount of packet loss in transmitting the plurality of packets including the one or more duplicates of the first packet to the destination computing node;
determining, based on the second amount of packet loss, a second duplication pattern for a second plurality of packets to be transmitted to the destination computing node as a second sequence of packets, wherein the second duplication pattern includes (i) a second number of candidate packets of the second sequence of packets to be duplicated, (ii) a second number of duplicates of each of the second number of candidate packets, and (iii) a second non-zero spacing between duplicates of each of the second number of candidate packets;
duplicating a second packet within the second plurality of packets according to the second duplication pattern, including generating one or more duplicates of the second packet and spacing the one or more duplicates within the plurality of packets according to the second non-zero spacing; and
transmitting the second plurality of packets including the one or more duplicates of the second packet to the destination computing node.
8. A non-transitory computer-readable medium storing instructions, which when executed by a processing device, cause the processing device to perform a method comprising:
determining an amount of packet loss in transmitting packets to a destination computing node;
determining, based on the amount of packet loss, a duplication pattern for a plurality of packets to be transmitted to the destination computing node as a sequence of packets, wherein the duplication pattern includes (i) a number of candidate packets of the sequence of packets to be duplicated, (ii) a number of duplicates of each of the candidate packets, and (iii) a non-zero spacing between duplicates of each of the candidate packets;
duplicating a first packet within the plurality of packets according to the duplication pattern, including generating one or more duplicates of the first packet and spacing the one or more duplicates within the plurality of packets according to the non-zero spacing;
transmitting the plurality of packets including the one or more duplicates of the first packet to the destination computing node;
receiving a first acknowledgement from the destination computing node; and
in accordance with a determination that the first acknowledgement is directed to a duplicate of the first packet:
in accordance with a determination that a second acknowledgement has yet to be received for each of a first one or more packets within the plurality of packets transmitted prior to the duplicate of the first packet, retransmitting the first one or more packets to the destination computing node.
9. The non-transitory computer-readable medium of claim 8 , further comprising calculating the number of candidate packets as a predetermined percentage of the amount of packet loss.
10. The non-transitory computer-readable medium of claim 9 , further comprising, in accordance with a determination that the amount of packet loss is more than a predetermined threshold, capping the number of candidate packets at a predetermined maximum value.
11. The non-transitory computer-readable medium of claim 8 , further comprising determining the non-zero spacing based on a predetermined minimum number of non-duplicated packets to be transmitted between duplicated packets.
12. The non-transitory computer-readable medium of claim 8 , further comprising duplicating the first packet after the sequence of packets.
13. The non-transitory computer-readable medium of claim 8 , the duplication pattern further includes a respective number of the duplicates of each of the candidate packets.
14. The non-transitory computer-readable medium of claim 8 , further comprising:
subsequent to retransmitting the first one or more packets to the destination computing node, determining a second amount of packet loss in transmitting the plurality of packets including the one or more duplicates of the first packet to the destination computing node;
determining, based on the second amount of packet loss, a second duplication pattern for a second plurality of packets to be transmitted to the destination computing node as a second sequence of packets, wherein the second duplication pattern includes (i) a second number of candidate packets of the second sequence of packets to be duplicated, (ii) a second number of duplicates of each of the second number of candidate packets, and (iii) a second non-zero spacing between duplicates of each of the second number of candidate packets;
duplicating a second packet within the second plurality of packets according to the second duplication pattern, including generating one or more duplicates of the second packet and spacing the one or more duplicates within the plurality of packets according to the second spacing; and
transmitting the second plurality of packets including the one or more duplicates of the second packet to the destination computing node.
15. An apparatus comprising:
a processing device; and
a memory coupled to the processing device, the memory storing instructions which, when executed by the processing device, cause the apparatus to:
determine an amount of packet loss in transmitting packets to a destination computing node;
determine, based on the amount of packet loss, a duplication pattern for a plurality of packets to be transmitted to the destination computing node as a sequence of packets, wherein the duplication pattern includes (i) a number of candidate packets of the sequence of packets to be duplicated, (ii) a number of duplicates of each of the candidate packets, and (iii) a non-zero spacing between duplicates of each of the candidate packets;
duplicate a first packet within the plurality of packets according to the duplication pattern, including generating one or more duplicates of the first packet and spacing the one or more duplicates within the plurality of packets according to the non-zero spacing;
transmit the plurality of packets including the one or more duplicates of the first packet to the destination computing node;
receive a first acknowledgement from the destination computing node; and
in accordance with a determination that the first acknowledgement is directed to a duplicate of the first packet:
in accordance with a determination that a second acknowledgement has yet to be received for each of a first one or more packets within the plurality of packets transmitted prior to the duplicate of the first packet, retransmit the first one or more packets to the destination computing node.
16. The apparatus of claim 15 , wherein the number of candidate packets is calculated as a predetermined percentage of the amount of packet loss.
17. The apparatus of claim 16 , wherein, in accordance with a determination that the amount of packet loss is more than a predetermined threshold, the number of candidate packets is capped at a predetermined maximum value.
18. The apparatus of claim 15 , wherein the non-zero spacing is calculated based on a predetermined minimum number of non-duplicated packets to be transmitted between duplicated packets.
19. The apparatus of claim 15 , wherein the first packet is duplicated after the sequence of packets.
20. The apparatus of claim 15 , wherein the duplication pattern further includes a respective number of the duplicates of each of the candidate packets.