Title: TCP Westwood: Experiments over Large Pipes
1TCP Westwood Experiments over Large Pipes
- Cesar Marcondes
- Anders Persson
- Prof. M.Y. Sanadidi
- Prof. Mario Gerla
- NRL Network Research Lab
- UCLA
2Background
- TCP NewReno is challenged on large pipes
- Slow convergence to full utilization
- Not intended to handle non-congestion packet loss
- Large Pipes performance criteria
- Utilization
- Stability
- Fast Ramp Up to Cruising Speed from Slow start
- Fairness under differing RTTs
- Friendliness to NewReno
- Alternatives include HS TCP, FAST, TCPW
- Goal of this study Measurements of TCPW, FAST
and HS TCP over large pipes
3TCPW
- Goal high utilization, fairness, and
friendliness over large leaky dynamic pipes - Sender side only estimation of Eligible Rate
Estimate (ERE) - Estimation takes into account congestion level,
capacity of the bottleneck, achieved rate - Exponential filtering to time average estimates
and avoid network conditions instability - ERE is used to
- (1) set congestion window after packet loss
- (2) repeatedly reset ssthresh to reach cruising
speed fast from slow start
4TCPW ABSE
RE Sampling Packet train, fair estimate
under congestion, underestimates under random
loss
- BE Sampling
- Packet pair,
- effective under random loss, overestimates
under congestion
Under Congestion
Under No Congestion
- To obtain ERE adapt the sample interval Tk
according to congestion level - Congestion level is similar to that in Vegas
Expected Rate-Achieved Rate
5Experiments Environment
NewReno Sender
Gigabit link
Internet2
NewReno Receiver (Alabama)
Gigabit link
UCLA Gigabit Switch
Advanced TCP Sender
- (Powerful Machines)
- CPU Xeon 3.06GHz
- Cache 512 L2/ 1MB L3
- Intel 1000PRO
- PCI-X BUS 133MHz
NewReno Receiver (Caltech)
6UCLA Internet2 Link Traffic
Other UCLA Users in Background
Our Experiments Traffic
7Test Methodology
- Automated Scripts
- Scheduled by Unix crontab
- Automatically reinitiate the O.S. with each
protocol and conduct new measurements - Linux FAST, HS-TCP and NewReno
- FreeBSD TCPW
- Sender/Receiver buffer is set to 2 MB to enable
high utilization of Gbps links - Iperf traffic generation, TCPdump, Nistnet
emulator
8Benchmark Tests
- Case Study I
- UCLA-Alabama (155 Mbps, 64 msec)
- Case Study II
- UCLA-CalTech (1 Gbps, 4msec)
- Group of 10 successive night time runs for each
test - Throughput, fairness, friendliness
- Artificial non-congestion loss (PER 0.1 to 0.5)
9Case Study I UCLAAlabama
NewReno Sender
Internet2 (Gigabit)
ATM Atlanta Alabama
NewReno Receiver (Alabama)
Advanced TCP Sender
155Mbps ATM Link Bottleneck Link as measured by
PathRate And confirmed later by the network admin
10Throughput
UCLA-Alabama
- Convergence to cruising speed varies among
protocols - High deviation among multiple runs in HSTCP and
NewReno - HSTCP deviations decrease over time (as the AIMD
behavior changes)
11UCLA-Alabama
12Transfer Completion Times
UCLA-Alabama
- On average
- TCPW and FAST 0 to 100 MB in 5.8 Sec!
- HSTCP 0 to 100 MB in 7.5 Sec!
- NewReno 0 to 100 MB in 11 Sec!
13Friendliness
UCLA-Alabama
14TCP FAST Preliminary Analysis
UCLA-Alabama
RTT Variation over Time as Observed by TCPdump
Outstanding Window as Observed by TCPdump
15Random Loss Emulation
UCLA-Alabama
UCLA Alabama
NewReno Receiver (Alabama)
Advanced TCP Sender
Nistnet Network Emulator
- Induced non-congestion packet loss in emulator
(PER 0.1 up to 0.5) - TCPW throughput much higher than all other schemes
16Random Loss Emulation (Results)
UCLA-Alabama
17Case Study II UCLACalTech
NewReno Sender (UCLA)
Internet2 (Gigabit)
Advanced TCP Sender (UCLA)
TCP Receiver (CalTech)
1 Gbps 4 ms
18Throughput
UCLA-CalTech
- TCP NewReno starts-up really high since it relies
in the cached threshold and the feedback is
really fast - Cached Slow Start Threshold versus Adaptive
Start-Up (Pros and Cons) - Westwood is delayed by its own Stability Filter
- Stability-based Filter dampens estimates in
proportion to the variance of observation
19UCLA-CalTech
20TCP Westwood Stability Filter versus Fixed Gain
Filter
UCLA-CalTech
- Sample Estimations vary a lot due to NIC
coalescing and OS issues at Gigabit/s. - As variability increases, stability filter relies
on a more stable moving average filter - Solution Use a fixed gain instead of an adaptive
when we know we are dealing with Gbps range
speeds - TCPW ramp up as HS-TCP and FAST
21TCPW Start-Up using Fixed Exponential Average
UCLA-CalTech
22Friendliness
UCLA-CalTech
23Conclusions
- TCPW and FAST performed equally well in terms of
average throughput - All Advanced TCP protocols have an excellent
intra-protocol fairness - Friendliness
- FAST appears to suffer a synchronization problem
- Under non-congestion error scenario, TCPW shows
greater robustness - At Gigabit speed, measurements could be messed up
by Interrupt Coalescing and other HW/Kernel
bottlenecks, affecting moving average filters
24Future Work
- New algorithm that is Interrupt Coalescence-Aware
for Gbps environment - New Agile and Stable Filter
- Improve the Automated TCP Test Tool (Benchmark
and New Tests)
25Thanks
- Netlab CalTech
- Xiaoyan Hong CS / Alabama Univ.