On the Robustness of Soft-State Protocols

1
On the Robustness of Soft-State Protocols

• John Lui, CUHK
• Vishal Misra, Columbia U.
• Dan Rubenstein, Columbia U.

2
State

• To operate correctly, network protocols require
  that communicating nodes share state, e.g.,
• Connection is active
• The largest sequence received was
• Q In networks with a lossy/unpredictable control
  channel, how is state information kept consistent
  across nodes?

3
Keeping State Consistent

• Two very different approaches / philosophies /
  mantras to how the signaling is performed
• Hard-state The Telephony Philosophy?
• Soft-state The Internet Philosophy Clark89
• The difference
• Easy to describe philosophically
• Hard to define precisely

4
Soft-state signaling
Sender
Receiver
Signaling plane
Communication plane

• Best effort signaling
• Refresh timer state needs periodic refresh
• State only removed by time-out
• Failure to communicate ? go to safe (default)
  state

5
Soft-state signaling
Sender
Receiver
Signaling plane
Communication plane

• Best effort signaling
• Refresh timer state needs periodic refresh
• State only removed by time-out
• Failure to communicate ? go to safe (default)
  state

6
Hard-state signaling
Sender
Receiver
X
Signaling plane
Communication plane

• State is explicitly added and removed
• Assumes very reliable communication channel
• Failure to communicate ? special recovery
  procedure

7
So Why is Soft State Design Better?

• Some common responses
• Its more robust
• To what? Packet loss? High delays?
• Its better at handling really bizarre network
  conditions
• Like what? Really high loss rates? Really high
  delays?
• Recovery is part of soft states normal operating
  process (no separate recovery operations needed)
• So what?

8
Prior work examining Soft State

• Raman,McCanne 99
• Queueing model of SS signaling system
• Showed SS/HS hybrid improves protocol performance
• Ji et al 03
• Performance comparison between SS, HS, and SS/HS
  hybrids
• Conclusion Hard State beats Soft State, but
  hybrid SS/HS protocols are best

So Why is Soft State Design Better?
9
Whats Wrong with Traditional Performance
Evaluations

• Tradition Given some network conditions, design
  the best protocol.

Input Conditions
Protocol Parameters
10
Whats Wrong with Traditional Performance
Evaluations

• Tradition Given some network conditions, design
  the best protocol.

Input Conditions
Output Best Solution
Protocol Parameters
11
The Traditional Conclusion

• For any network condition, hard state protocols
  can be configured for that condition to
  out-perform their soft state counterparts

12
A more practical performance evaluation

• Dont really know what the conditions will be
  when configuring the protocol

Input Conditions
Output (Best?) Solution
Protocol Parameters
Is Hard State best in this setting?
13
Performance-Oriented View of Protocol Designer
Intuition
Hard State Protocol

• Suppose protocols are tuned to operate most
  efficiently under normal conditions
• Claim HS performance worsens more rapidly than
  SS as conditions vary from norm

bad
Normal Operating Regime
Soft State Protocol
Performance
good
Network Condition
14
Our Comparison Study

• We choose 3 network scenarios
• DoS Attack
• Correlated, Lossy Feedback Channel
• Broadcast Communication Environment
• For each scenario
• Pick a HS and SS protocol used in the scenario
• Choose protocol parameters (timeout lengths,
  attempts) to work well for expected network
  conditions
• Vary the network conditions
• Watch how the protocol performs (w/o rechoosing
  protocol parameters!!)

15
A Generic Signaling Protocol Model

• L Lifetime that a state should exist
• R Refresh interval
• T Timeout interval (e.g., 3R for SS many
  protocols)
• p Channel loss probability
• K1 , K2 , etc. Various Costs (described later)
  

16
Refresh Cost
Sender
Receiver
Signaling plane
Communication plane
Cost to keep state consistent
17
(Re)Initialization Cost
Sender
Receiver
Signaling plane
Communication plane
Cost to recover from accidental timeout
of drops pL/R, Cost K2 pL/R
18
Stale state cost
Sender
Receiver
Signaling plane
Communication plane
Cost of enacting an actual timeout
Stale state lifetime R, Cost K3 pR
19
Total Cost
C(R) K2 p L/R K1 L/R K3 p R EC(R) K2 p
EL/R K1 EL/R K3 p R
What is the optimal R to minimize total cost?
20
Optimal R implications

• K2 ,K1 large ? Performance emphasis
• Fewer refresh pings, bad to tear down state
  accidentally
• K3 large ? Robustness emphasis
• Bad to miss tearing down state
• Higher R, Harder the protocol, Lower R,
  Softer the protocol

21
Cost Comparison
Results match previous robustness intuition
22
Resource Blocking (DoS) Attacks

• Good Traffic uses and releases resource
• Attacker doesnt release resource until timeout

Hard state more susceptible to attacks
23
Correlated, Lossy Feedback Channel

• Client connects to a server
• If loss rate from server too high, client chooses
  to disconnect
• Soft State receiver stops sending refresh
  messages
• Hard State receiver tries to push a disconnect
  message through the lossy channel
• Channel losses (in both directions) are equal

24
The Hard-State Dilemma
STOP!
STOP!
STOP!
Feedback loop Inability to terminate induces
greater losses, making it more difficult to
terminate
25
Results of Markov Model Formulation
As session expected lifetime (1/µ) decreases, HS
zombie sessions grow large
Soft State has many fewer zombie sessions
26
Robust Multicast Feedback

• Scenario sender broadcasts transmission as long
  as some receiver listening
• Q How does sender know if a receiver is
  listening?

27
Hard State Approach

• Each interested receiver explicitly notifies
  sender of join and leave

R
R
S
R
28
Soft State Approach

• Some receiver must ping sender about interest
  within time period T or broadcast stops
• receiver pings randomly delayed and broadcast so
  other receivers can suppress their pings
• propagation delays can induce multiple pings per
  interval

R
R
S
R
T
T
T
T
29
Optimized Versions

• Prefix-matching methods Bolot93 can be used to
  reduce receiver communication costs
• Hard-state used to choose a leader
• Soft-sate used to reduce feedback rate

30
Heavy Arrival Rate Comparison
? arrival rate of interested clients
Soft State designs exhibit better scalability
with large ? for both versions of polling
protocols
31
Heavy Departure Rate Comparison
µ departure rate of interested clients
Soft State designs exhibit better scalability
with large µ for both versions of polling
protocols
32
Conclusions

• Hard state protocols can often outperform soft
  state protocols when network conditions are known
• What makes soft state better design is its
  ability to provide acceptable performance over
  a larger variety of network conditions