{"draft":"draft-ietf-tcpm-rack-15","doc_id":"RFC8985","title":"The RACK-TLP Loss Detection Algorithm for TCP","authors":["Y. Cheng","N. Cardwell","N. Dukkipati","P. Jha"],"format":["HTML","TEXT","PDF","XML"],"page_count":"29","pub_status":"PROPOSED STANDARD","status":"PROPOSED STANDARD","source":"TCP Maintenance and Minor Extensions","abstract":"This document presents the RACK-TLP loss detection algorithm for TCP.\r\nRACK-TLP uses per-segment transmit timestamps and selective\r\nacknowledgments (SACKs) and has two parts. Recent Acknowledgment\r\n(RACK) starts fast recovery quickly using time-based inferences\r\nderived from acknowledgment (ACK) feedback, and Tail Loss Probe (TLP)\r\nleverages RACK and sends a probe packet to trigger ACK feedback to\r\navoid retransmission timeout (RTO) events. Compared to the widely\r\nused duplicate acknowledgment (DupAck) threshold approach, RACK-TLP\r\ndetects losses more efficiently when there are application-limited\r\nflights of data, lost retransmissions, or data packet reordering\r\nevents. It is intended to be an alternative to the DupAck threshold\r\napproach.","pub_date":"February 2021","keywords":["TCP","Loss Recovery","Reordering"],"obsoletes":[],"obsoleted_by":[],"updates":[],"updated_by":[],"see_also":[],"doi":"10.17487\/RFC8985","errata_url":null}