Distributed Snapshots:

1
Distributed Snapshots

• Non-blocking checkpoint coordination protocol

Next Uncoordinated Chkpnt
2
Uncoordinated

• Processes take chkpnt independently
• Domino Effect!

Next Coordinated Blocking Chkpnt
3
Coordinated Blocking

• Processes are coordinated to form a consistent
  global state, and

okay, channels flushed
Ready!
Go!
initiator

p1

p2

p3
Next Coordinated Blocking Chkpnt (cont)
4
Coordinated Blocking (cont)

• Advantage
• Always consistent
• No Domino Effect
• Less storage overhead
• Disadvantage
• Large latency to chkpnt!

Next Coordinated Non-blocking Chkpnt
5
Coordinated Non-blocking

• Processes are coordinated, but
• Do we really need to block ?

!
K. Mani Chandy
Leslie Lamport
!
Next Global-state Recording Algorithm
6
Global-state Recording Alg.
Distributed snapshots determining global
states of distributed systems, K. Mani
Chandy and Leslie Lamport

• Step 1 process states
• Step 2 channel states
• Step 3 end of the algorithm

Next Model of Distributed System
7
Model of Distributed System

• Processes
• Channels directed, FIFO, error-free

Next Step 1, process states
8
Step 1 process states

• Initiator
• Save its local state
• Send marker tokens on all outgoing edges
• All other processes
• On receiving the first marker on any incoming
  edges,
• Save state, and propagate markers on all outgoing
  edges
• Resume execution.
• Further markers will be eaten up.

Next Example
9

• Example

p
x
x
q
x
x
x
r
Next Proof
10

• Proof

Let us assume that a message m exists, and it
makes our cut inconsistent.
p
m
q
Next Proof (cont)
11

• Proof(cont)

Incomplete page
p
m
x1

• x1 is the 1st marker
• for process q

q
x2
p
m
(2) x1 is not the 1st marker for process q
x1
q
x2
Contradict the assumption.
Next Step 2, channel states
12
Step 2 channel states
p
In-flight messages
q

• Sent along the channel before the senders
  chkpnt
• Received along the channel after the receivers
  chkpnt

Next Example
13

• Example

(1) p is receiving messages
(2) p has just saved its state
r
r
s
s
q
q
x
x
7
7
x
x
8
8
5
5
x
3
6
6
2
1
4
4
p
p
x
x
u
u
t
t
Next Example (cont)
14

• Example(cont)

ps chkpnt triggered by a marker from q
r
s
x
q
x
7
1
2
3
5
4
6
7
8
p
x
8
5
x
x
3
6
q
2
1
4
x
x
x
p
r
x
s
u
t
x
Next Algorithm (revised)
15
Algorithm (revised)

• Initiator
• Save its local state
• Send marker tokens on all outgoing edges
• All other processes
• On receiving the first marker on any incoming
  edges,
• Save state, and propagate markers on all outgoing
  edges
• Resume execution, but also save incoming messages
  until a marker arrives through the channel
• Guarantees a consistent global state!

Next Step 3, end of the algorithm
16
Step 3 end of the algorithm

• Did every process save its state and in-flight
  messages?

• direct channel to the initiator?
• spanning tree?
• General solution?

Next References
17
References

• Distributed snapshots determining global
• States of distributed systems,
• K. Mani Chandy and Leslie Lamport