1. © British Telecommunications plc 1 Network Performance Isolation in Data Centres using ConEx Congestion Policing draft-briscoe-conex-policing-01 draft-briscoe-conex-data-centre-02 draft-briscoe-conex-policing-01 draft-briscoe-conex-data-centre-02 Bob Briscoe Chief Researcher, BT IRTF DCLC Jul 2014 Bob Briscoe’s work is part-funded by the European Community under its Seventh Framework Programme through the Trilogy 2 project (ICT-317756)

2. © British Telecommunications plc 2 purpose of talk work proposal for the data centre latency control r-g –data centre queuing delay control –designed for global scope (inter-data-centre,... Inter-net) –this talk: adds first step: intra-data centre without any new protocols started in the IETF congestion exposure (ConEx) w-g generalised for initial deployment without ConEx –and even without ECN end-to-end –now even without ECN on switches (in slides, not draft)

3. © British Telecommunications plc 3 Network Performance Isolation in Data Centres An important problem –isolating between tenants, or departments –virtualisation isolates CPU / memory / storage –but network and I/O system is highly multiplexed & distributed SDN-based (edge) capacity partitioning* –configuration churn: nightmare at scale –poor use of capacity edge-based weighted round robin (or WFQ) –More common –but biases towards heavy hitters (no concept of time) bit-rate time bit-rate time * every problem in computer science can be solved by not thinking then hiding the resulting mess under a layer of abstraction

4. © British Telecommunications plc 4 Outline Design – First Step edge bottlenecks by capacity design Edge policing like Diffserv –but congestion policing (per guest) isolation within FIFO queue no config on switches hosts switches VM sender VM receiver congestion policer guest OS hypervisor switching w

5. © British Telecommunications plc 5 bottleneck congestion policer foreach pkt { i = classify_user(pkt) d i += w i *(t now -t i ) //fill t i = t now d i -= s * p //drain if (d i <0) {drop(pkt)} } s : packet size p : drop prob of AQM in a well-provisioned link, policer rarely intervenes but whenever needed, it limits queue growth network policer incoming packet stream meter w1w1 w2w2 wiwi … d i (t) cici congestion token bucket Y FIFO buffer p(t) AQM … outgoing packet stream

6. © British Telecommunications plc 6 wiwi d i (t) cici kw i C AQM p(t) policer main congestion token bucket meter congestion burst limiter police if either bucket empties actually each bucket needs to be two buckets to limit bursts of congestion similar code –except 2 token buckets... if (d i1 <0 || d i2 <0) {drop(pkt)}

7. © British Telecommunications plc 7 performance isolation outcome WRR or WFQ congestion policer –with unequal traffic loads congestion policer –treats equal traffic loads equivalently to WRR time rate time rate time rate

8. © British Telecommunications plc 8 Outline Design edge and core queue control Edge policing like Diffserv –but congestion policing (per-guest) Hose model intra-class isolation in all FIFO queues FIFO ECN marking on L3 switches no other config on switches hosts switches VM sender VM receiver TEP / congestion policer TEP / audit guest OS hypervisor switching w

9. © British Telecommunications plc 9 trusted path congestion feedback Initial deployment –all under control of infrastructure admin ECN on guest hosts: optional –ECN enabled across tunnel ConEx on guest hosts: optional –any ConEx-enabled packet doesn’t require tunnel feedback details – see spare slide or draft infrastructure policer transport sender transport receiver TEP feedback between tunnel endpoints Non-ConEx packets infrastructure policeraudit transport sender transport receiver ConEx packets

10. © British Telecommunications plc 10 Features of Solution Network performance isolation between tenants No loss of LAN-like multiplexing benefits work-conserving Zero (tenant-related) switch configuration No change to existing switch implementations Weighted performance differentiation Simplest possible contract per-tenant network-wide allowance tenant can freely move VMs around without changing allowance sender constraint, but with transferable allowance Transport-Agnostic Extensible to wide-area and inter-data-centre interconnect

11. © British Telecommunications plc 11 call for interest implementation in hypervisors evaluation

12. Network Performance Isolation in Data Centres using congestion policing draft-briscoe-conex-policing-01 draft-briscoe-conex-data-centre-02 draft-briscoe-conex-policing-01 draft-briscoe-conex-data-centre-02 Q&A & spare slides

13. © British Telecommunications plc 13 measuring contribution to congestion = bytes weighted by congestion level = bytes dropped (or ECN-marked) = ‘congestion-volume’ as simple to measure as volume bit-rate time congestion time 10GB 0.01% congestion 1MB 1% congestion 1MB 300MB 100MB 3MB 1% 0.01%

14. unilateral deployment technique for data centre operator for e2e transports that don’t support ECN, the operator can: 1.at encap: alter 00 to 10 in outer 2.at interior buffers: turn on ECN defers any drops until egress audit just before egress can see packets to be dropped 14 exploits: widespread edge-edge tunnels in multi-tenant DCs to isolate forwarding a side-effect of standard tunnelling (IP-in-IP or any ECN link encap) DSDS ECNECN ingress egress DSDS ECNECN DSDS ECNECN DSDS ECNECN DSDS ECNECN DSDS ECNECN E congested network element DSDS ECNECN DSDS ECNECN E 12 00  10 10  11 11  drop A audit A pol- icer for e2e transports that don’t support ConEx, the operator can create its own trusted feedback: 3.at decap: only for Not-ConEx packets, feedback aggregate congestion marking counters: CE outer, Not-ECT inner = loss CE outer, ECT inner = ECN exporter collector IPFIX meter 3