802.1Qbb – Priority-based Flow Control
The full title of this PAR is “IEEE Standard for Local and Metropolitan Area Networks—Virtual Bridged Local Area Networks – Amendment: Priority-based Flow Control.”.
This standard specifies protocols, procedures and managed objects that enable flow control per traffic class on IEEE 802 full-duplex links. Data Center Bridging networks (bridges and end nodes) are characterized by limited bandwidth-delay product and limited hop-count. Traffic class is identified by the VLAN tag priority values. Priority-based flow control is intended to eliminate frame loss due to congestion. This is achieved by a mechanism similar to the IEEE 802.3x PAUSE, but operating on individual priorities. This mechanism, in conjunction with other Data Center Bridging technologies, enables support for higher layer protocols that are highly loss sensitive while not affecting the operation of traditional LAN protocols utilizing other priorities. In addition, PFC complements Congestion Notification in Data Center Bridging networks. Operation of priority-based flow control is limited to a domain controlled by a Data Center Bridging control protocol that controls the application of Priority-based Flow Control, Enhanced Transmission Selection, and Congestion Notification.
Data Center Bridging networks employ higher layer protocols that depend on the delivery of data frames without frame loss due to congestion. These protocols were designed for an underlying transport that approaches lossless behavior and therefore do not include appropriate response to frame loss due to congestion (e.g. back-off, slow restart, etc.). This amendment enables multiple data center networks, including those serving loss sensitive protocols (e.g. inter-processor communitcation, storage, etc.), to be converged onto an IEEE 802 network.
There is significant customer interest and market opportunity for 802 LANs as a converged Layer 2 solution in high-speed short-range networks such as data centers, backplane fabrics, single and multi-chassis interconnects, computing clusters, and storage networks. These environments currently use Layer 2 networks that do not discard packets due to congestion (e.g., Fibre Channel, InfiniBand). This project will bring comparable frame loss characteristics to 802 LANs in Data Center Bridging environments. This in conjunction with the other Data Center Bridging technologies enable converged networks. Use of a converged network will realize operational and equipment cost benefits.
Current Status
Status | PAR approved March 27th, 2008; Standard approved Jun 16th, 2011 |
Editor | Claudio DeSanti |
Archive
Presentations
Date | Document |
---|---|
27 Mar 2008 | Presentation bb-pelissier-pfc-proposal-0308.pdf |
31 Mar 2008 | Presentation bb-thaler-par-0308.pdf |
07 Apr 2008 | Presentation bb-pelissier-pfc-proposal-0408.pdf |
07 Apr 2008 | Presentation bb-pelissier-pfc-proposal-0408v2.pdf |
21 Apr 2008 | Presentation bb-pelissier-pfc-proposal-0408v3.pdf |
08 May 2008 | Presentation bb-pelissier-pfc-proposal-0508.pdf |
15 May 2008 | Presentation bb-pelissier-pfc-proposal-dcb-comments-0508.pdf |
16 Jul 2008 | Presentation bb-thaler-frame-issues.pdf |
17 Jul 2008 | Presentation bb-pelissier-pfc-mac-control-0708.pdf |
08 Nov 2008 | Presentation bb-pelissier-convergence-proposal-1108.pdf |
11 Nov 2008 | Presentation bb-cds-draft-0-2-status-1108.pdf |
16 Apr 2009 | Presentation bb-pelissier-pfc-defense-0409.pdf |
17 May 2009 | Presentation bb-ghanwani-pfc-untagged-0509-v1.pdf |
09 Sep 2009 | Presentation bb-cds-state-diagrams-0909-v01.pdf |
11 Sep 2009 | Presentation bb-cds-state-diagrams-0909-v02.pdf |