algorithms and scheduling techniques for heterogeneous...
TRANSCRIPT
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Algorithms and scheduling techniques forheterogeneous platforms
Yves Robert
Ecole Normale Superieure de [email protected]
http://graal.ens-lyon.fr/∼yrobert
joint work withOlivier Beaumont, Anne Benoit, Larry Carter, Henri Casanova,
Jeanne Ferrante, Arnaud Legrand, Loris Marchal, Frederic Vivien
Tokyo, July 2007
[email protected] POP’07 Algorithms and scheduling techniques 1/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Evolution of parallel machines
From (good old) parallel architectures . . .
[email protected] POP’07 Algorithms and scheduling techniques 2/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Evolution of parallel machines
. . . to heterogeneous clusters . . .
[email protected] POP’07 Algorithms and scheduling techniques 2/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Evolution of parallel machines
. . . and to large-scale grid platforms?
[email protected] POP’07 Algorithms and scheduling techniques 2/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Evolution of parallel machines
. . . and to large-scale grid platforms?
Parallel algorithm design and scheduling were already difficult taskswith homogeneous machines
[email protected] POP’07 Algorithms and scheduling techniques 2/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Evolution of parallel machines
. . . and to large-scale grid platforms?
Parallel algorithm design and scheduling were already difficult taskswith homogeneous machinesOn heterogeneous platforms, it gets worse
[email protected] POP’07 Algorithms and scheduling techniques 2/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
New platforms, new problems, new solutions
Target platforms: Large-scale heterogenous platforms(networks of workstations, clusters, collections of clusters, grids, ...)
New problems
Heterogeneity of processors (CPU power, memory)
Heterogeneity of communication links
Irregularity of interconnection networks
Non-dedicated platforms
Need to adapt algorithms and scheduling strategies: new objectivefunctions, new models
[email protected] POP’07 Algorithms and scheduling techniques 3/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
New platforms, new problems, new solutions
Target platforms: Large-scale heterogenous platforms(networks of workstations, clusters, collections of clusters, grids, ...)
New problems
Heterogeneity of processors (CPU power, memory)
Heterogeneity of communication links
Irregularity of interconnection networks
Non-dedicated platforms
Need to adapt algorithms and scheduling strategies: new objectivefunctions, new models
[email protected] POP’07 Algorithms and scheduling techniques 3/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Outline
1 Parallel algorithms for heterogeneous clusters
2 Background on traditional scheduling
3 Packet routing
4 Master-worker on heterogeneous platforms
5 Limitations & going further
6 Conclusion
[email protected] POP’07 Algorithms and scheduling techniques 4/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Outline
1 Parallel algorithms for heterogeneous clusters
2 Background on traditional scheduling
3 Packet routing
4 Master-worker on heterogeneous platforms
5 Limitations & going further
6 Conclusion
[email protected] POP’07 Algorithms and scheduling techniques 5/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Independent chunks
B independent equal-size tasks • • • • • • • • ••p processors P1, P2, . . ., Pp
wi = time for Pi to process a task •
Intuition: load of Pi proportional to its speed 1/wi
Assign ni tasks to Pi
Objective: minimize Texe = max∑pi=1 ni=B
(ni × wi )
[email protected] POP’07 Algorithms and scheduling techniques 6/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Dynamic programming
With 3 processors: w1 = 3, w2 = 5, and w3 = 8
P1 • • • • • • • • •P2 N N N N N M M M M M N N N N NP3
Task n1 n2 n3 Texe Selected proc.0 0 0 0 11 1 0 0 3 22 1 1 0 5 13 2 1 0 6 34 2 1 1 8 15 3 1 1 9 26 3 2 1 10 17 4 2 1 12 18 5 2 1 15 29 5 3 1 15 310 5 3 2 16
[email protected] POP’07 Algorithms and scheduling techniques 7/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Static versus dynamic
Greedy (demand-driven) would have done a perfect job
Would even be better (possible variations in processor speeds)
Static assignment required useless thinking /
[email protected] POP’07 Algorithms and scheduling techniques 8/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Static versus dynamic
Greedy (demand-driven) would have done a perfect job
Would even be better (possible variations in processor speeds)
Static assignment required useless thinking /
[email protected] POP’07 Algorithms and scheduling techniques 8/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Coping with dependences
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
x x xx x x
j
i
x x xx x x
x x xx x x
N
N
1
2
x x xx x x
x x xx x x
T2,3
A simple finite difference problem
Iteration space: 2D rectangle of size N1 × N2
Dependences between tiles
(10
),
(01
)
[email protected] POP’07 Algorithms and scheduling techniques 9/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (1/3)
Use column-wise allocation to enhance locality
...
6
5 → 6
4 → 5 → 6
3 → 4 → 5
2 → 3 → 4
1 → 2 → 3 . . .
Stepwise execution
[email protected] POP’07 Algorithms and scheduling techniques 10/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (1/3)
Use column-wise allocation to enhance locality
...
6
5 → 6
4 → 5 → 6
3 → 4 → 5
2 → 3 → 4
1 → 2 → 3 . . .
Stepwise execution
[email protected] POP’07 Algorithms and scheduling techniques 10/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (1/3)
Use column-wise allocation to enhance locality
...
6
5 → 6
4 → 5 → 6
3 → 4 → 5
2 → 3 → 4
1 → 2 → 3 . . .
Stepwise execution
[email protected] POP’07 Algorithms and scheduling techniques 10/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (1/3)
Use column-wise allocation to enhance locality
...
6
5 → 6
4 → 5 → 6
3 → 4 → 5
2 → 3 → 4
1 → 2 → 3 . . .
Stepwise execution
[email protected] POP’07 Algorithms and scheduling techniques 10/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (1/3)
Use column-wise allocation to enhance locality
...
6
5 → 6
4 → 5 → 6
3 → 4 → 5
2 → 3 → 4
1 → 2 → 3 . . .
Stepwise execution
[email protected] POP’07 Algorithms and scheduling techniques 10/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (1/3)
Use column-wise allocation to enhance locality
...
6
5 → 6
4 → 5 → 6
3 → 4 → 5
2 → 3 → 4
1 → 2 → 3 . . .
Stepwise execution
[email protected] POP’07 Algorithms and scheduling techniques 10/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (1/3)
Use column-wise allocation to enhance locality
...
6
5 → 6
4 → 5 → 6
3 → 4 → 5
2 → 3 → 4
1 → 2 → 3 . . .
Stepwise execution
[email protected] POP’07 Algorithms and scheduling techniques 10/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (2/3)
With column-wise allocation,
Topt ≈N1 × N2∑p
i=11wi
.
Greedy (demand-driven) allocation ⇒ slowdown ?!Execution progresses at the pace of the slowest processor /
[email protected] POP’07 Algorithms and scheduling techniques 11/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (2/3)
With column-wise allocation,
Topt ≈N1 × N2∑p
i=11wi
.
Greedy (demand-driven) allocation ⇒ slowdown ?!Execution progresses at the pace of the slowest processor /
[email protected] POP’07 Algorithms and scheduling techniques 11/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Allocation strategy (3/3)
With 3 processors, w1 = 3, w2 = 5, and w3 = 8:
...
33
w =3
w =5 1
2
w =8P
P
P1
2
i
j
Texe ≈8
3N1N2 ≈ 2.67 N1N2
Topt ≈120
79N1N2 ≈ 1.52 N1N2
[email protected] POP’07 Algorithms and scheduling techniques 12/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Periodic static allocation (1/2)
With 3 processors, w1 = 3, w2 = 5, and w3 = 8:
...
33
w =3
w =5 1
2
w =8P
P
P1
2
i
j
Assigning blocks of B = 10 columns, Texe ≈ 1.6 N1N2
[email protected] POP’07 Algorithms and scheduling techniques 13/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Periodic static allocation (2/2)
L = lcm(w1,w2, . . . ,wp)Example: L = lcm(3, 5, 8) = 120
P1 receives first n1 = L/w1 columns, P2 next n2 = L/w2
columns, and so on
Period: block of B = n1 + n2 + . . . + np contiguous columnsExample: B = n1 + n2 + n3 = 40 + 24 + 15 = 79
Change schedule:- Sort processors so that n1w1 ≤ n2w2 ≤ . . . ≤ npwp
- Process horizontally within blocks
Optimal ,
[email protected] POP’07 Algorithms and scheduling techniques 14/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Lesson learnt?
With different-speed processors . . .. . . we need to think (design static schedules)
. . . but implementation may remain dynamic ,
Example: demand-driven allocation of blocks of adequate size
. . . well, in some cases it gets truly complicated /
[email protected] POP’07 Algorithms and scheduling techniques 15/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Lesson learnt?
With different-speed processors . . .. . . we need to think (design static schedules)
. . . but implementation may remain dynamic ,
Example: demand-driven allocation of blocks of adequate size
. . . well, in some cases it gets truly complicated /
[email protected] POP’07 Algorithms and scheduling techniques 15/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Lesson learnt?
With different-speed processors . . .. . . we need to think (design static schedules)
. . . but implementation may remain dynamic ,
Example: demand-driven allocation of blocks of adequate size
. . . well, in some cases it gets truly complicated /
[email protected] POP’07 Algorithms and scheduling techniques 15/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Outline
1 Parallel algorithms for heterogeneous clusters
2 Background on traditional scheduling
3 Packet routing
4 Master-worker on heterogeneous platforms
5 Limitations & going further
6 Conclusion
[email protected] POP’07 Algorithms and scheduling techniques 16/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Traditional scheduling – Framework
Application = DAG G = (T ,E ,w)
T = set of tasksE = dependence constraintsw(T ) = computational cost of task T (execution time)c(T ,T ′) = communication cost (data sent from T to T ′)
Platform
Set of p identical processors
Schedule
σ(T ) = date to begin execution of task Talloc(T ) = processor assigned to it
[email protected] POP’07 Algorithms and scheduling techniques 17/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Traditional scheduling – Constraints
w(T’)
time
w(T)
comm(T,T’)
σ(T ) + w(T )σ(T )
T T’
σ(T ′)
Data dependences If (T ,T ′) ∈ E then
if alloc(T ) = alloc(T ′) then σ(T ) + w(T ) ≤ σ(T ′)if alloc(T ) 6= alloc(T ′) then σ(T ) + w(T ) + c(T ,T ′) ≤ σ(T ′)
Resource constraints
alloc(T ) = alloc(T ′) ⇒(σ(T ) + w(T ) ≤ σ(T ′)) or (σ(T ′) + w(T ′) ≤ σ(T ))
[email protected] POP’07 Algorithms and scheduling techniques 18/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Traditional scheduling – Objective functions
Makespan or total execution time
MS(σ) = maxT∈T
(σ(T ) + w(T ))
Other classical objectives:
Sum of completion timesWith release dates: max flow (response time), or sum flowFairness oriented: max stretch, or sum stretch
[email protected] POP’07 Algorithms and scheduling techniques 19/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Traditional scheduling – About the model
Simple but OK for computational resources
No CPU sharing, even in models with preemptionAt most one task running per processor at any time-step
Very crude for network resources
Unlimited number of simultaneous sends/receives per processorNo contention → unbounded bandwidth on any linkFully connected interconnection graph (clique)
In fact, model assumes infinite network capacity
[email protected] POP’07 Algorithms and scheduling techniques 20/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Makespan minimization
NP-hardness
Pb(p) NP-complete for independent tasks and nocommunications(E = ∅, p = 2 and c = 0)Pb(p) NP-complete for UET-UCT graphs (w = c = 1)
Approximation algorithms
Without communications, list scheduling is a(2− 1
p )-approximationWith communications, result extends to coarse-grain graphsWith communications, no λ-approximation in general
[email protected] POP’07 Algorithms and scheduling techniques 21/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
List scheduling – Without communications
Initialization:
Compute priority level of all tasks
Priority queue = list of free tasks (tasks without predecessors)sorted by priority
While there remain tasks to execute:
Add new free tasks, if any, to the queue.
If there are q available processors and r tasks in the queue,remove first min(q, r) tasks from the queue and execute them
Priority level
Use critical path: longest path from the task to an exit node
Computed recursively by a bottom-up traversal of the graph
[email protected] POP’07 Algorithms and scheduling techniques 22/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
List scheduling – Without communications
Initialization:
Compute priority level of all tasks
Priority queue = list of free tasks (tasks without predecessors)sorted by priority
While there remain tasks to execute:
Add new free tasks, if any, to the queue.
If there are q available processors and r tasks in the queue,remove first min(q, r) tasks from the queue and execute them
Priority level
Use critical path: longest path from the task to an exit node
Computed recursively by a bottom-up traversal of the graph
[email protected] POP’07 Algorithms and scheduling techniques 22/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
List scheduling – With communications (1/2)
Priority level
Use pessimistic critical path: include all edge costs in theweightComputed recursively by a bottom-up traversal of the graph
MCP Modified Critical Path
Assign free task with highest priority to best processorBest processor = finishes execution first, given already takenscheduling decisionsFree tasks may not be ready for execution (communicationdelays)May explore inserting the task in empty slots of scheduleComplexity O(|V | log |V |+ (|E |+ |V |)p)
[email protected] POP’07 Algorithms and scheduling techniques 23/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
List scheduling – With communications (2/2)
EFT Earliest Finish Time
Dynamically recompute priorities of free tasksSelect free task that finishes execution first (on bestprocessor), given already taken scheduling decisionsHigher complexity O(|V |3p)May miss “urgent” tasks on critical path
Other approaches
Two-step: clustering + load balancing- DSC Dominant Sequence Clustering O((|V |+ |E |) log |V |)- LLB List-based Load Balancing O(C log C + |V |) (C numberof clusters generated by DSC)Low-cost: FCP Fast Critical Path- Maintain constant-size sorted list of free tasks:- Best processor = first idle or the one sending last message- Low complexity O(|V | log p + |E |)
[email protected] POP’07 Algorithms and scheduling techniques 24/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Extending the model to heterogeneous clusters
Task graph with n tasks T1, . . . ,Tn.
Platform with p heterogeneous processors P1, . . . ,Pp.
Computation costs:- wiq = execution time of Ti on Pq
- wi =∑p
q=1 wiq
p average execution time of Ti
- particular case: consistent tasks wiq = wi × γq
Communication costs:- data(i , j): data volume for edge eij : Ti → Tj
- vqr : communication time for unit-size message from Pq toPr (zero if q = r)- com(i , j , q, r) = data(i , j)× vqr communication time from Ti
executed on Pq to Pj executed on Pr
- comij = data(i , j)×∑
1≤q,r≤p,q 6=r vqr
p(p−1) average communicationcost for edge eij : Ti → Tj
[email protected] POP’07 Algorithms and scheduling techniques 25/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Rewriting constraints
Dependences For eij : Ti → Tj , q = alloc(Ti ) and r = alloc(Tj):
σ(Ti ) + wiq + com(i , j , q, r) ≤ σ(Tj)
Resources If q = alloc(Ti ) = alloc(Tj), then
(σ(Ti ) + wiq ≤ σ(Tj)) or (σ(Tj) + wjq ≤ σ(Ti ))
Makespanmax
1≤i≤n
(σ(Ti ) + wi ,alloc(Ti )
)
[email protected] POP’07 Algorithms and scheduling techniques 26/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
HEFT: Heterogeneous Earliest Finish Time
Priority level:
rank(Ti ) = wi + maxTj∈Succ(Ti )
(comij + rank(Tj)),
where Succ(T ) is the set of successors of TRecursive computation by bottom-up traversal of the graph
Allocation
For current task Ti , determine best processor Pq:minimize σ(Ti ) + wiq
Enforce constraints related to communication costsInsertion scheduling: look for t = σ(Ti ) s.t. Pq is availableduring interval [t, t + wiq[
Complexity: same as MCP without/with insertion
[email protected] POP’07 Algorithms and scheduling techniques 27/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
What’s wrong?
, Nothing (still may need to map a DAG onto a platform!)
/ Absurd communication model:complicated: many parameters to instantiatewhile not realistic (clique + no contention)
/ Wrong metric: need to relax makespan minimizationobjective
[email protected] POP’07 Algorithms and scheduling techniques 28/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
What’s wrong?
, Nothing (still may need to map a DAG onto a platform!)
/ Absurd communication model:complicated: many parameters to instantiatewhile not realistic (clique + no contention)
/ Wrong metric: need to relax makespan minimizationobjective
[email protected] POP’07 Algorithms and scheduling techniques 28/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
What’s wrong?
, Nothing (still may need to map a DAG onto a platform!)
/ Absurd communication model:complicated: many parameters to instantiatewhile not realistic (clique + no contention)
/ Wrong metric: need to relax makespan minimizationobjective
[email protected] POP’07 Algorithms and scheduling techniques 28/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Outline
1 Parallel algorithms for heterogeneous clusters
2 Background on traditional scheduling
3 Packet routing
4 Master-worker on heterogeneous platforms
5 Limitations & going further
6 Conclusion
[email protected] POP’07 Algorithms and scheduling techniques 29/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Problem
A
E
G
C
H
DF
B
Routing sets of messages from sources to destinations
Paths not fixed a priori
Packets of same message may follow different paths
[email protected] POP’07 Algorithms and scheduling techniques 30/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Hypotheses
A
E
G
C
H
DF
B
A packet crosses an edge within one time-step
At any time-step, at most one packet crosses an edge
[email protected] POP’07 Algorithms and scheduling techniques 31/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Hypotheses
A
E
G
C
H
DF
B
A packet crosses an edge within one time-step
At any time-step, at most one packet crosses an edge
Scheduling: for each time-step, decide which packet crosses anygiven edge
[email protected] POP’07 Algorithms and scheduling techniques 31/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Notation
k
i
l
j
nk ,l
nk ,li ,j
nk,l : total number of packets to be routed from k to l
nk,li ,j : total number of packets routed from k to l and crossing
edge (i , j)
[email protected] POP’07 Algorithms and scheduling techniques 32/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Lower bound
Congestion Ci ,j of edge (i , j)= total number of packets that cross (i , j)
Ci ,j =∑
(k,l)|nk,l>0
nk,li ,j Cmax = maxi ,j Ci ,j
Cmax lower bound on schedule makespanC ∗ ≥ Cmax
⇒ “Fluidified” solution in Cmax?
[email protected] POP’07 Algorithms and scheduling techniques 33/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Lower bound
Congestion Ci ,j of edge (i , j)= total number of packets that cross (i , j)
Ci ,j =∑
(k,l)|nk,l>0
nk,li ,j Cmax = maxi ,j Ci ,j
Cmax lower bound on schedule makespanC ∗ ≥ Cmax
⇒ “Fluidified” solution in Cmax?
[email protected] POP’07 Algorithms and scheduling techniques 33/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Equations (1/2)
A
E
G
C
H
DF
B
A
E
G
B
Initialization (packets leave node k):∑
j |(k,j)∈A
nk,lk,j = nk,l
Reception (packets reach node l):∑
i |(i ,l)∈A
nk,li ,l = nk,l
Conservation law (crossing intermediate node i):∑i |(i ,j)∈A
nk,li ,j =
∑i |(j ,i)∈A
nk,lj ,i ∀(k, l), j 6= k, j 6= l
[email protected] POP’07 Algorithms and scheduling techniques 34/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Equations (1/2)
A
E
G
C
H
DF
B
G
H
D
Initialization (packets leave node k):∑
j |(k,j)∈A
nk,lk,j = nk,l
Reception (packets reach node l):∑
i |(i ,l)∈A
nk,li ,l = nk,l
Conservation law (crossing intermediate node i):∑i |(i ,j)∈A
nk,li ,j =
∑i |(j ,i)∈A
nk,lj ,i ∀(k, l), j 6= k, j 6= l
[email protected] POP’07 Algorithms and scheduling techniques 34/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Equations (1/2)
G G
Initialization (packets leave node k):∑
j |(k,j)∈A
nk,lk,j = nk,l
Reception (packets reach node l):∑
i |(i ,l)∈A
nk,li ,l = nk,l
Conservation law (crossing intermediate node i):∑i |(i ,j)∈A
nk,li ,j =
∑i |(j ,i)∈A
nk,lj ,i ∀(k, l), j 6= k, j 6= l
[email protected] POP’07 Algorithms and scheduling techniques 34/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Equations (2/2)
CongestionCi ,j =
∑(k,l)|nk,l>0 nk,l
i ,j
Objective function
Cmax ≥ Ci ,j , ∀i , j
Minimize Cmax
Linear program in rational numbers: polynomial-time solution. Inpractice use GLPK, Maple, Mupad . . .
[email protected] POP’07 Algorithms and scheduling techniques 35/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Equations (2/2)
CongestionCi ,j =
∑(k,l)|nk,l>0 nk,l
i ,j
Objective function
Cmax ≥ Ci ,j , ∀i , j
Minimize Cmax
Linear program in rational numbers: polynomial-time solution. Inpractice use GLPK, Maple, Mupad . . .
[email protected] POP’07 Algorithms and scheduling techniques 35/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Routing algorithm
Compute optimal solution Cmax, nk,li ,j of previous linear
program
Periodic schedule:
Define Ω =√
Cmax
Use⌈
Cmax
Ω
⌉periods of length Ω
During each period, edge (i , j) forwards (at most)
mk,li,j =
⌊nk,l
i,j Ω
Cmax
⌋
packets that go from k to l
Clean-up: sequentially process residual packets inside network
[email protected] POP’07 Algorithms and scheduling techniques 36/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Performance
Schedule is feasible
Schedule is asymptotically optimal:
Cmax ≤ C ∗ ≤ Cmax + O(√
Cmax)
[email protected] POP’07 Algorithms and scheduling techniques 37/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Why does it work?
Relaxation of objective function
Rational number of packets in LP formulation
Periods long enough so that rounding down to integernumbers has negligible impact
Periods numerous enough so that loss in first and last periodshas negligible impact
Periodic schedule, described in compact form
[email protected] POP’07 Algorithms and scheduling techniques 38/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Outline
1 Parallel algorithms for heterogeneous clusters
2 Background on traditional scheduling
3 Packet routing
4 Master-worker on heterogeneous platforms
5 Limitations & going further
6 Conclusion
[email protected] POP’07 Algorithms and scheduling techniques 39/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Master-worker tasking: framework
Heterogeneous resources
Processors of different speedsCommunication links with various bandwidths
Large number of independent tasks to process
Tasks are atomicTasks have same size
Single data repository
One master initially holds data for all tasksSeveral workers arranged along a star, a tree ora general graph
[email protected] POP’07 Algorithms and scheduling techniques 40/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Application examples
Monte Carlo methods
SETI@home
Factoring large numbers
Searching for Mersenne primes
Particle detection at CERN (LHC@home)
... and many others: see BOINC at http://boinc.berkeley.edu
[email protected] POP’07 Algorithms and scheduling techniques 41/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Makespan vs. steady state
Two-different problems
Makespan Maximize total number of tasks processed within atime-bound
Steady state Determine periodic task allocation which maximizestotal throughput
[email protected] POP’07 Algorithms and scheduling techniques 42/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
!" #" $ #" % # $ &' %'!" #" $ #" % # $ &' %'
( ) *,+ -./0 1 23 45 6
78 9:; < 9=> ?@
AB CDE FE G B D E H I AJ K LM NO PQ RS T U M UV Q N T
W N UX RY X O V M UX NQ O X TZ M N ZQ Y [\ UX UQ Q
[email protected] POP’07 Algorithms and scheduling techniques 43/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
Time for computingTime for computing
one task in Cone task in C
Time for sending Time for sending
one task from A to Bone task from A to B
A is the root of the tree;A is the root of the tree;
all tasks start at Aall tasks start at A
[email protected] POP’07 Algorithms and scheduling techniques 43/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
A computeA computeA sendA send
B receiveB receiveB computeB compute
C receiveC receiveC computeC compute
C sendC send
D receiveD receiveD computeD compute
11 22 33
[email protected] POP’07 Algorithms and scheduling techniques 43/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
A computeA computeA sendA send
B receiveB receiveB computeB compute
C receiveC receiveC computeC compute
C sendC send
D receiveD receiveD computeD compute
11 22 33
[email protected] POP’07 Algorithms and scheduling techniques 43/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
A computeA computeA sendA send
B receiveB receiveB computeB compute
C receiveC receiveC computeC compute
C sendC send
D receiveD receiveD computeD compute
11 22 33
[email protected] POP’07 Algorithms and scheduling techniques 43/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
A computeA computeA sendA send
B receiveB receiveB computeB compute
C receiveC receiveC computeC compute
C sendC send
D receiveD receiveD computeD compute
11 22 33
[email protected] POP’07 Algorithms and scheduling techniques 43/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
A computeA computeA sendA send
B receiveB receiveB computeB compute
C receiveC receiveC computeC compute
C sendC send
D receiveD receiveD computeD compute
11 22 33
[email protected] POP’07 Algorithms and scheduling techniques 43/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
A computeA computeA sendA send
B receiveB receiveB computeB compute
C receiveC receiveC computeC compute
C sendC send
D receiveD receiveD computeD compute
11 22 33
StartupStartupRepeatedRepeated
patternpatternCleanClean--upup
SteadySteady--state: 7 tasks every 6 time unitsstate: 7 tasks every 6 time units
[email protected] POP’07 Algorithms and scheduling techniques 43/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Solution for star-shaped platforms
Communication links between master and workers havedifferent bandwidths
Workers have different computing power
[email protected] POP’07 Algorithms and scheduling techniques 44/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Rule of the game
M
P1 P2 Pi Pp
w1 w2 wi wp
ci
cpc1
c2
Master sends tasks to workers sequentially, and withoutpreemption
Full computation/communication overlap for each worker
Worker Pi receives a task in ci time-units
Worker Pi processes a task in wi time-units
[email protected] POP’07 Algorithms and scheduling techniques 45/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Equations
M
P1 P2 Pi Pp
w1 w2 wi wp
ci
cpc1
c2
Worker Pi executes αi tasks per time-unit
Computations: αiwi ≤ 1
Communications:∑
i αici ≤ 1
Objective: maximize throughput
ρ =∑
i
αi
[email protected] POP’07 Algorithms and scheduling techniques 45/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Solution
Faster-communicating workers first: c1 ≤ c2 ≤ . . .
Make full use of first q workers, where q largest index s.t.
q∑i=1
ci
wi≤ 1
Make partial use of next worker Pq+1
Discard other workers
Bandwidth-centric strategy- Delegate work to the fastest communicating workers- It doesn’t matter if these workers are computing slowly- Slow workers will not contribute much to overall throughput
[email protected] POP’07 Algorithms and scheduling techniques 46/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
Fully active
210
201
3 6 1 1 1
M
3
Discarded
Tasks Communication Computation6 tasks to P1 6c1 = 6 6w1 = 183 tasks to P2 3c2 = 6 3w2 = 182 tasks to P3 2c3 = 6 2w3 = 2
11 tasks every 18 time-units (ρ = 11/18 ≈ 0.6)
[email protected] POP’07 Algorithms and scheduling techniques 47/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Example
Fully active
210
201
3 6 1 1 1
M
3
Discarded
, Compare to purely greedy (demand-driven) strategy!5 tasks every 36 time-units (ρ = 5/36 ≈ 0.14)
Even if resources are cheap and abundant,resource selection is key to performance
[email protected] POP’07 Algorithms and scheduling techniques 47/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Extension to trees
Fully used node
Partially used node
Idle node
1432
1 1 1
3
0
2
555
26666
5 5
9
10
12
12
Resource selection based on local information (children)
[email protected] POP’07 Algorithms and scheduling techniques 48/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Does this really work?
Can we deal with arbitrary platforms (including cycles)?
Can we deal with return messages?
In fact, can we deal with more complex applications (arbitrarycollections of DAGs)?
[email protected] POP’07 Algorithms and scheduling techniques 49/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Does this really work?
Can we deal with arbitrary platforms (including cycles)? Yes
Can we deal with return messages?
In fact, can we deal with more complex applications (arbitrarycollections of DAGs)?
[email protected] POP’07 Algorithms and scheduling techniques 49/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Does this really work?
Can we deal with arbitrary platforms (including cycles)? Yes
Can we deal with return messages?
In fact, can we deal with more complex applications (arbitrarycollections of DAGs)?
[email protected] POP’07 Algorithms and scheduling techniques 49/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Does this really work?
Can we deal with arbitrary platforms (including cycles)? Yes
Can we deal with return messages? Yes
In fact, can we deal with more complex applications (arbitrarycollections of DAGs)?
[email protected] POP’07 Algorithms and scheduling techniques 49/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Does this really work?
Can we deal with arbitrary platforms (including cycles)? Yes
Can we deal with return messages? Yes
In fact, can we deal with more complex applications (arbitrarycollections of DAGs)?
[email protected] POP’07 Algorithms and scheduling techniques 49/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Does this really work?
Can we deal with arbitrary platforms (including cycles)? Yes
Can we deal with return messages? Yes
In fact, can we deal with more complex applications (arbitrarycollections of DAGs)? Yes, I mean, almost!
[email protected] POP’07 Algorithms and scheduling techniques 49/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
LP formulation still works well . . .
Tm
file emnPj
Pi
Pk
wi
cik
cji
Tn
Conservation law
∀m, n∑
j
sent(Pj → Pi , emn) + executed(Pi ,Tm)
= executed(Pi ,Tn) +∑k
sent(Pi → Pk , emn)
Computations∑m
executed(Pi ,Tm)× flops(Tm)× wi ≤ 1
Outgoing communications∑m,n
∑j
sent(Pj → Pi , emn)× bytes(emn)× cij ≤ 1
[email protected] POP’07 Algorithms and scheduling techniques 50/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
. . . but schedule reconstruction is harder
P1 → P2
P2 → P1
P1 → P3
P3 → P1
P2 → P4
P4 → P2
P3 → P4
P4 → P3
P2 → P3
P3 → P2
P4
P3
P2
P1
χ3 χ4 χ1 χ2 χ3 χ4 χ1 χ2 χ3 χ4 χ1 χ2 χ3 χ4 χ1 χ2
0 40 80 120 160
A5A4A3A2A1
, Actual (cyclic) schedule obtained in polynomial time
, Asymptotic optimality
/ A couple of practical problems (large period, # buffers)
/ No local scheduling policy
[email protected] POP’07 Algorithms and scheduling techniques 51/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
The beauty of steady-state scheduling
Rationale Maximize throughput (total load executed perperiod)
Simplicity Relaxation of makespan minimization problem
Ignore initialization and clean-up phasesPrecise ordering of tasks/messages not neededCharacterize resource activity per time-unit:- which (rational) fraction of time is spentcomputing for which application?- which (rational) fraction of time is spentreceiving from or sending to which neighbor?
Efficiency Optimal throughput ⇒ optimal schedule (up to aconstant number of tasks)
Periodic schedule, described in compact form⇒ compiling a loop instead of a DAG!
[email protected] POP’07 Algorithms and scheduling techniques 52/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Lesson learnt?
Resource selection is mandatory
. . . implementation may still be dynamic,provided that static allocation is enforced by scheduler
Example: demand-driven assignment of enrolled workers
[email protected] POP’07 Algorithms and scheduling techniques 53/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Lesson learnt?
Resource selection is mandatory
. . . implementation may still be dynamic,provided that static allocation is enforced by scheduler
Example: demand-driven assignment of enrolled workers
[email protected] POP’07 Algorithms and scheduling techniques 53/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Outline
1 Parallel algorithms for heterogeneous clusters
2 Background on traditional scheduling
3 Packet routing
4 Master-worker on heterogeneous platforms
5 Limitations & going further
6 Conclusion
[email protected] POP’07 Algorithms and scheduling techniques 54/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Good news and bad news
, One-port model: first step towards designing realisticscheduling heuristics
, Steady-state circumvents complexity of schedulingproblems. . . while deriving efficient (often asympotically optimal)scheduling algorithms
/ Need to acquire a good knowledge of the platform graph
/ Need to run extensive experiments or simulations
[email protected] POP’07 Algorithms and scheduling techniques 55/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Good news and bad news
, One-port model: first step towards designing realisticscheduling heuristics
, Steady-state circumvents complexity of schedulingproblems. . . while deriving efficient (often asympotically optimal)scheduling algorithms
/ Need to acquire a good knowledge of the platform graph
/ Need to run extensive experiments or simulations
[email protected] POP’07 Algorithms and scheduling techniques 55/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Knowledge of the platform graph
For regular problems, the structure of the task graph (nodesand edges) only depends upon the application, not upon thetarget platform
Problems arise from weights, i.e. the estimation of executionand communication times
Classical answer: “use the past to predict the future”
Divide scheduling into phases, during which machine andnetwork parameters are collected (with NWS)⇒ This information guides scheduling decisions for next phase
Moving from heterogeneous clusters to computational gridscauses further problems (even discovering the characteristicsof the surrounding computing resources may prove a difficulttask)
[email protected] POP’07 Algorithms and scheduling techniques 56/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Knowledge of the platform graph
For regular problems, the structure of the task graph (nodesand edges) only depends upon the application, not upon thetarget platform
Problems arise from weights, i.e. the estimation of executionand communication times
Classical answer: “use the past to predict the future”
Divide scheduling into phases, during which machine andnetwork parameters are collected (with NWS)⇒ This information guides scheduling decisions for next phase
Moving from heterogeneous clusters to computational gridscauses further problems (even discovering the characteristicsof the surrounding computing resources may prove a difficulttask)
[email protected] POP’07 Algorithms and scheduling techniques 56/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Knowledge of the platform graph
For regular problems, the structure of the task graph (nodesand edges) only depends upon the application, not upon thetarget platform
Problems arise from weights, i.e. the estimation of executionand communication times
Classical answer: “use the past to predict the future”
Divide scheduling into phases, during which machine andnetwork parameters are collected (with NWS)⇒ This information guides scheduling decisions for next phase
Moving from heterogeneous clusters to computational gridscauses further problems (even discovering the characteristicsof the surrounding computing resources may prove a difficulttask)
[email protected] POP’07 Algorithms and scheduling techniques 56/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Experiments versus simulations
Real experiments difficult to drive (genuine instability ofnon-dedicated platforms)
Simulations ensure reproducibility of measured data
Key issue: run simulations against a realistic environment
Trace-based simulation: record platform parameters today,and simulate the algorithms tomorrow, against recorded data
Use SIMGRID, an event-driven simulation toolkit
[email protected] POP’07 Algorithms and scheduling techniques 57/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Experiments versus simulations
Real experiments difficult to drive (genuine instability ofnon-dedicated platforms)
Simulations ensure reproducibility of measured data
Key issue: run simulations against a realistic environment
Trace-based simulation: record platform parameters today,and simulate the algorithms tomorrow, against recorded data
Use SIMGRID, an event-driven simulation toolkit
[email protected] POP’07 Algorithms and scheduling techniques 57/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
SIMGRID traces
server #1
client #1
client #2
client #3
server #2
router
switch
hub
Internet
CPU availability
Network bandwidth
Transient Failure X
See http://simgrid.gforge.inria.fr/
[email protected] POP’07 Algorithms and scheduling techniques 58/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Across physical links
Network = directed graph P = (V ,E )
P0
P1
P3
P2
time
recv 2,3P3
T2,3(L)link e2,3
send2,3P2
General case: affine model (includes latencies)
Common variant: sending and receiving processors busyduring whole transfer
[email protected] POP’07 Algorithms and scheduling techniques 59/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Across physical links
Network = directed graph P = (V ,E )
P0
P1
P3
P2
time
r2,3
r2,3 · LP3
α2,3
β2,3 · Llink e2,3
s2,3 · Ls2,3P2
General case: affine model (includes latencies)
Common variant: sending and receiving processors busyduring whole transfer
[email protected] POP’07 Algorithms and scheduling techniques 59/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Across physical links
Network = directed graph P = (V ,E )
P0
P1
P3
P2
time
recv 2,3P3
T2,3(L)link e2,3
send2,3P2
General case: affine model (includes latencies)
Common variant: sending and receiving processors busyduring whole transfer
[email protected] POP’07 Algorithms and scheduling techniques 59/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Multi-port
Bar-Noy, Guha, Naor, Schieber:occupation time of sender Pu independent of target Pv
time
recv vPv
Tu,v(L)link eu,v
senduPu
not fully multi-port model, but allows for starting a new transfer
from Pu without waiting for previous one to finish
[email protected] POP’07 Algorithms and scheduling techniques 60/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
One-port
Bhat, Raghavendra and Prasanna:same parameters for sender Pu, link eu,v and receiver Pv
time
ru,v · Lru,vPv
βu,v · Lαu,vlink eu,v
su,v · Lsu,vPu
two flavors:- bidirectional: simultaneous send and receive transfers allowed
- unidirectional: only one send or receive transfer at a given
time-step
[email protected] POP’07 Algorithms and scheduling techniques 61/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Store & Forward, WormHole, TCP
How to model a file transfer along a path?
[email protected] POP’07 Algorithms and scheduling techniques 62/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Store & Forward, WormHole, TCP
How to model a file transfer along a path?
S
l1
l3
l2
[email protected] POP’07 Algorithms and scheduling techniques 62/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Store & Forward, WormHole, TCP
How to model a file transfer along a path?
S
l1
l3
l2
Store & Forward : bad model for contention
[email protected] POP’07 Algorithms and scheduling techniques 62/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Store & Forward, WormHole, TCP
How to model a file transfer along a path?
S
l1
l3
l2
[email protected] POP’07 Algorithms and scheduling techniques 62/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Store & Forward, WormHole, TCP
How to model a file transfer along a path?
pi ,j
s
S
l1
l3
l2
WormHole : computation intensive (packets), not that realistic
[email protected] POP’07 Algorithms and scheduling techniques 62/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Store & Forward, WormHole, TCP
How to model a file transfer along a path?
∀l ∈ L,∑
r∈R s.t. l∈r
ρr ≤ cl
Analytical model
[email protected] POP’07 Algorithms and scheduling techniques 62/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Store & Forward, WormHole, TCP
How to model a file transfer along a path?
∀l ∈ L,∑
r∈R s.t. l∈r
ρr ≤ cl
Max-Min Fairness maximize minr∈R
ρr
Proportional Fairness maximize∑r∈R
ρr log(ρr )
MCT minimization maximize minr∈R
1
ρr
TCP behavior Close to max-min.In SIMGRID: max-min + bound by 1/RTT
[email protected] POP’07 Algorithms and scheduling techniques 62/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Bandwidth sharing
Traditional assumption: Fair Sharing
Open i TCP connections, receive bw(i) bandwidth perconnection
bw(i) = bw(1)/i on a LAN
Experimental evidence → bw(i) = bw(1) on a WAN
Backbone links have so many connections that interferenceamong a few selected connections is negligible
Better model: bw(i) =bw(1)
1 + (i − 1).γ
γ = 1 for a perfect LAN, γ = 0 for a perfect WAN
[email protected] POP’07 Algorithms and scheduling techniques 63/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Bandwidth sharing
Traditional assumption: Fair Sharing
Open i TCP connections, receive bw(i) bandwidth perconnection
bw(i) = bw(1)/i on a LAN
Experimental evidence → bw(i) = bw(1) on a WAN
Backbone links have so many connections that interferenceamong a few selected connections is negligible
Better model: bw(i) =bw(1)
1 + (i − 1).γ
γ = 1 for a perfect LAN, γ = 0 for a perfect WAN
[email protected] POP’07 Algorithms and scheduling techniques 63/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Bandwidth sharing
Traditional assumption: Fair Sharing
Open i TCP connections, receive bw(i) bandwidth perconnection
bw(i) = bw(1)/i on a LAN
Experimental evidence → bw(i) = bw(1) on a WAN
Backbone links have so many connections that interferenceamong a few selected connections is negligible
Better model: bw(i) =bw(1)
1 + (i − 1).γ
γ = 1 for a perfect LAN, γ = 0 for a perfect WAN
[email protected] POP’07 Algorithms and scheduling techniques 63/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Sample large-scale platform
Primergy
Primergy
backbone link
router
front end
cluster
[email protected] POP’07 Algorithms and scheduling techniques 64/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
What topology?
Generated (GT-ITM, BRITE, etc.) or obtained frommonitoring?
Very complex (Layer 2 information)Not clear that a scheduling algorithm could exploit/know allthat information
Need a simple model that is
More accurate than traditional models (e.g., LAN links,fully-connected)Still amenable to analysis
[email protected] POP’07 Algorithms and scheduling techniques 65/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
What topology?
Generated (GT-ITM, BRITE, etc.) or obtained frommonitoring?
Very complex (Layer 2 information)Not clear that a scheduling algorithm could exploit/know allthat information
Need a simple model that is
More accurate than traditional models (e.g., LAN links,fully-connected)Still amenable to analysis
[email protected] POP’07 Algorithms and scheduling techniques 65/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
What topology? (cont’d)
Primergy
Primergy
uuγ = 0.5
Unknown topology
(complete graph)
[email protected] POP’07 Algorithms and scheduling techniques 66/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
What topology? (cont’d)
Primergy
Primergy
backbone link
router
front end
cluster
Hierarchy + BW sharing, but assume knowledge of
Routing
Backbone bandwidths
CPU speeds
[email protected] POP’07 Algorithms and scheduling techniques 67/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
A first trial
sk
sl
gk
gl
b3
b1
Lk,l
b2
C k
C kmaster
C krouter
C lrouter
C lmaster
C l
Clusters and backbone links
[email protected] POP’07 Algorithms and scheduling techniques 68/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
A first trial (cont’d)
sk
sl
gk
gl
b3
b1
Lk,l
b2
C k
C kmaster
C krouter
C lrouter
C lmaster
C l
Clusters
K clusters C k , 1 ≤ k ≤ K
C kmaster front-end processor
C krouter router to external world
sk cumulated speed of C k
gk bandwidth of the LAN link (γ = 1) from C kmaster to C k
router
[email protected] POP’07 Algorithms and scheduling techniques 69/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
A first trial (cont’d)
sk
sl
gk
gl
b3
b1
Lk,l
b2
C k
C kmaster
C krouter
C lrouter
C lmaster
C l
Network
Set R of routers and B of backbone links li
bw(li ) bandwidth available for a new connection
max-connect(li ) max. number of connections that can beopened
Fixed routing: path Lk,l of backbones from C krouter to C l
router
[email protected] POP’07 Algorithms and scheduling techniques 69/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Scheduling multiple applications
Large-scale platforms not likely to be exploited in dedicatedmode/single application
Investigate scenarios in which multiple applications aresimultaneously executed on the platform⇒ competition for CPU and network resources
[email protected] POP’07 Algorithms and scheduling techniques 70/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Target problem
Large complex platform: several clusters and backbone links
One (divisible load) application running on each cluster
Which fraction of the job to delegate to other clusters?
Applications have different communication-to-computationratios
How to ensure fair scheduling and good resource utilization?
[email protected] POP’07 Algorithms and scheduling techniques 71/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Linear program
Maximize mink
αkπk
,
under the constraints
(1a) ∀C k ,∑
l
αk,l = αk
(1b) ∀C k ,∑
l
αl ,k .τl ≤ sk
(1c) ∀C k ,∑l 6=k
αk,l .δk +∑j 6=k
αj ,k .δj ≤ gk
(1d) ∀li ,∑
li∈Lk,l
βk,l ≤ max-connect(li )
(1e) ∀k, l , αk,l .δk ≤ βk,l × gk,l
(1f) ∀k, l , αk,l ≥ 0
(1g) ∀k, l , βk,l ∈ N
(1)
[email protected] POP’07 Algorithms and scheduling techniques 72/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Approach
Solution to rational linear problem as comparator/upperbound
Several heuristics, greedy and LP-based
Use Tiers as topology generator, and then SIMGRID
[email protected] POP’07 Algorithms and scheduling techniques 73/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Methodology
-1000
0
1000
2000
3000
4000
5000
-1000 0 1000 2000 3000 4000 5000
Ver
tical
Dis
tanc
e
Horizontal Distance Number of nodes: 641, Number of links:934
Original Network
WANMANLAN
500
1000
1500
2000
2500
3000
3500
4000
0 500 1000 1500 2000 2500
Ver
tical
Dis
tanc
e
Horizontal Distance Number of nodes: 41, Number of links:45
Pruned Network WANMANLAN
Sample full and pruned Tiers topology
distribution
K 5, 7, . . . , 90log(bw(lk)), log(gk) normal (mean= log(2000), std=log(10))sk uniform, 1000 — 10000max-connect, δk , τk , πk uniform, 1 — 10
Platform parameters used in simulation
[email protected] POP’07 Algorithms and scheduling techniques 74/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Hints for implementation
Participants sharing resources in a Virtual Organization
Centralized broker managing applications and resources
Broker gathers all parameters of LP program
Priority factors
Various policies and refinements possible⇒ e.g. fixed number of connections per application
[email protected] POP’07 Algorithms and scheduling techniques 75/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Outline
1 Parallel algorithms for heterogeneous clusters
2 Background on traditional scheduling
3 Packet routing
4 Master-worker on heterogeneous platforms
5 Limitations & going further
6 Conclusion
[email protected] POP’07 Algorithms and scheduling techniques 76/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Key advantages of steady-state scheduling
Simplicity
From local equations to global behaviorThroughput characterized from activity variables
Efficiency
Periodic schedule, described in compact formAsymptotic optimality
Adaptability
Record observed performance during currentperiod, and inject information to computeschedule for next periodReact on the fly to resource variations
[email protected] POP’07 Algorithms and scheduling techniques 77/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Open problems
Decentralized scheduling
From local strategies to provably good performance?Adapt Awerbuch-Leighton algorithm for multicommodityflows?
Concurrent scheduling
Several workflows competing for resourcesMulti-criteria and fairness?Adapt economic models and buzz-words (e.g., Nashequilibrium)?
[email protected] POP’07 Algorithms and scheduling techniques 78/ 79
Introduction Parallel algorithms for heterogeneous clusters Background on traditional scheduling Packet routing Master-worker on heterogeneous platforms Limitations & going further Conclusion
Scheduling for large-scale platforms
If platform is well identified and relatively stable, try to:(i) accurately model hierarchical structure(ii) design well-suited scheduling algorithms
If platform is not stable enough, or if it evolves too fast,dynamic schedulers are the only option
Otherwise, grab any opportunity to
inject static knowledge into dynamic schedulers
/ Is this opportunity a niche?, Does it encompass a wide range of applications?
[email protected] POP’07 Algorithms and scheduling techniques 79/ 79