Download presentation

Presentation is loading. Please wait.

Published byAliza Jolles Modified over 3 years ago

1
VARUN GUPTA Carnegie Mellon University 1 Partly based on joint work with: Anshul Gandhi Mor Harchol-Balter Mike Kozuch (CMU) (CMU) (Intel Research)

2
2 Demand (t) Time (t) ? Data Center Capacity PROBLEM: Want to turn servers OFF/ON to match (t)… … and also minimize setup penalties!

3
3 (t) # Busy servers # Jobs Mean Job Size = 1s Setup delay = 200s Busy power = 240W

4
4 (t) # Busy servers # Jobs Mean Job Size = 1s Setup delay = 200s Busy power = 240W

5
5 (t) # Busy servers # Jobs Mean Job Size = 1s Setup delay = 200s Busy power = 240W

6
6 (t) # Busy servers # Jobs Mean Job Size = 1s Setup delay = 200s Busy power = 240W WITH SETUP: E[Power] = 51.9 kW E[Response time] = 13s NO SETUP: E[Power] = 14.4 kW E[Response time] = 1s

7
7 (t) # Busy servers # Jobs Mean Job Size = 1s Setup delay = 200s Busy power = 240W Can we do better than ON/OFF?. Add inertia while turning servers OFF

8
8 Turn OFF a server after idle for t wait t wait independent of (t)! new arrival routed to the most recently busy (MRB) server arriving job turns a server ON if all servers busy

9
9 t wait timers

10
10 t wait timers

11
11 t wait timers

12
12 THEOREM: For Poisson arrivals, as the load, the number of ON servers is concentrated around --- Proof Intuition --- Step 1: An equivalent system view k jobs run on the “first” k servers Step 2: Analysis of idle periods of M/G/∞ Time from a k (k-1) to the next (k-1) k transition

13
13 Intuition for idle periods 1. # jobs Normal with mean and variance 2. 3. Events happen at rate 4. Mean idle period of server 1 2 3 MRB servers for any constant t wait ! In practice, we choose t wait to amortize setup cost

14
14 Mean Job Size = 1s Setup delay = 200s Busy power = 240W (t) # Busy+idle servers # Jobs ON/OFF DELAYEDOFF

15
15 Mean Job Size = 1s Setup delay = 200s Busy power = 240W (t) # Busy+idle servers # Jobs ON/OFF DELAYEDOFF

16
16 Mean Job Size = 1s Setup delay = 200s Busy power = 240W (t) # Busy+idle servers # Jobs ON/OFF DELAYEDOFF DELAYEDOFF: E[Power] = 18.9 kW E[Response time] = 1.002s NO SETUP: E[Power] = 14.4 kW E[Response time] = 1s.

17
17 Mean Job Size = 1s Setup delay = 200s Busy power = 240W (t) # Busy+idle servers # Jobs DELAYEDOFF SQ. ROOT W/ LOOKAHEAD (“optimal offline”)

18
18 Mean Job Size = 1s Setup delay = 200s Busy power = 240W (t) # Busy+idle servers # Jobs DELAYEDOFF SQ. ROOT W/ LOOKAHEAD (“optimal offline”)

19
19 Mean Job Size = 1s Setup delay = 200s Busy power = 240W (t) # Busy+idle servers # Jobs DELAYEDOFF DELAYEDOFF: E[Power] = 18.9 kW E[Response time] = 1.002s w/LOOKAHEAD: E[Power] = 15.8 kW E[Response time] = 1.036s SQ. ROOT W/ LOOKAHEAD (“optimal offline”)

20
1. Speed scaling algorithms 2. A simple proxy for MRB routing 3. Heterogeneous servers 4. Managing Virtual Machines in the Cloud 5. Wear-leveling/Performance tradeoffs 20

21
21 Power P(s) Speed (s) Q: Optimal speed to balance energy-performance? A: your favorite metric =

22
22 MRB requires a lot of state updates Proxy policy Assign static ranks to servers Route a new arrival to highest ranked idle server Almost the same performance as MRB + easier to implement than MRB + easy to extend (MRB) Rank 1 2 3 4 5 Rank- based

23
23 more efficient less efficient Assign ranks to servers based on efficiency (1 = most efficient) Route a new arrival to highest ranked idle server 1 2 3 4 5

24
24 2GB RAM 1GB

25
25 1 2 3 4 5 6 Split physical servers into “virtual” servers Assign static ranks to virtual servers Route a new VM request to highest ranked idle virtual server Turn the physical server OFF after each virtual server has idled for t wait

26
Problem: unpredictable demand and non-trivial setup costs DELAYEDOFF : A new traffic- oblivious capacity scaling scheme Extensions to real-world scenarios 26 Demand ? Time Max Capacity

27
27 THEOREM [MRB]: As the load, the number of ON servers is concentrated around THEOREM [Round-Robin]: As the load, for constant job sizes, the number of ON servers is N servers Arrival rate = λ

Similar presentations

OK

Do you have the Maths Factor?. Maths Can you beat this term’s Maths Challenge?

Do you have the Maths Factor?. Maths Can you beat this term’s Maths Challenge?

© 2017 SlidePlayer.com Inc.

All rights reserved.

Ads by Google

Ppt on conservation of wildlife and natural vegetation regions Ppt on microsoft dynamics nav erp Close reading elementary ppt on cells Ppt on automobile industry in india Ppt on world heritage day Download ppt on solar energy Ppt on famous indian entrepreneurship Download ppt on indus valley civilization pictures Download ppt on federalism in india Ppt on total parenteral nutrition procedure