Architecture of Computing Systems (Lecture Notes in Computer ...

Recommendations

Info

76 M. Bonn and H. Schmeck 2.2 Node Monitoring A central assumption during the design of JoSchKa was the heterogeneity of the nodes. They not only differ with respect to their hard- and software configuration, but also in their reliability characteristics. Some nodes process every job successfully but others break down in an unpredictable manner. They may also change their reliability or switch from unreliable to robust and vice versa. Imagine a PC in a student pool, which runs nonstop for many hours in the night, but gets rebooted very often at daytime, when students use this PC. To assess the behavior of the nodes a monitoring component was developed. The goal is to reach the ability to predict the behavior of the node in the near future [5]. Amongst others, the following values are monitored and calculated for each node, respectively: • �: The relative CPU power of the node compared to the others, ��. • ��: The average uptime in wall clock minutes, �� . • ��: The current uptime in wall clock minutes, �� . • �: A synthetic reliability index of the node, ��−1,1�. In the following, let ��,…,�� be the exponential weighted average, which processes the single input values �� considering their age. It is defined recursively: ��,…,�� =�� =�� 1 −�� with ��0,1� and �� =��. The �� are sorted by age, �� is the youngest. More formally, the values for a single node ��1, … , �� are calculated as follows: • �� = ∑ �� , where �� denotes the time span (in wall clock milliseconds) the �� node needs to process a fixed arithmetic-logical benchmark. This benchmark is done by the node in periodic intervals. • �� =�� ,…,��, where the �� are the nodes’ last �� uptimes. • �� =�� −��, where �� denotes the current wall clock time and �� denotes the starting wall clock time of the node. • �� =�� ,..,��, where the �� are reliability criteria for the last �� processed jobs. If a job is done successfully, then �� =1, if failed �� =−1. In practice, we chose �� =�� =10 and � =0.25. These data which are collected by the server for each single node can be summarized as follows: For each node, we now know, • how fast it can work compared to all other nodes (�), • how long it is available (on average) before it breaks down (��), • how long it is up since its last boot (��) and • how reliable it behaved when processing the last jobs (�). This knowledge is used by the different distribution strategies, which are described in the following section.
2.3 Basic Job Distribution Strategies The JoSchKa System 77 As pointed out, the JoSchKa system is not designed for specialized clusters (although it can be run there too, of course), but to distribute jobs within a pool of strongly varying nodes. The nodes differ in performance but mainly they differ in their reliability. It is possible that a node works perfectly reliably for many days, but gets suddenly rebooted repeatedly and unpredictably. Then, long running jobs can’t be processed any more on such a node. On the other hand, it would be a waste of valuable uptime, if a reliable node just processes jobs which are finished after a few minutes. So, if the distribution system disposes of many jobs which belong to many different users, it’s obvious to distribute the jobs in an intelligent way. The goal is to minimize the breakdown-caused waste of processing time and to be concerned with fairness to the users, too. No user should feel disadvantaged. At first, we describe the basic distribution strategies, in the following section (simulation) they are evaluated. JoSchKa uses various (combined) strategies, for lack of space we only show two of them here: The simplest algorithm and the one which normally is used in practice. (a) Balanced (fair) distribution: Every user gets the same share of nodes to process his jobs. (b) Uptime-based distribution: The decision, which job is selected for a polling agent is based on the agent’s uptime behavior. Normally, the scheduler does not differentiate between individual users, but single JobTypes. A job type is a user-defined subset of his jobs. So the different needs of an individual user who has his jobs grouped to probably more than one job type can be satisfied much better. If the number of known agents is less than the number of types, the scheduler does differentiate only the users, the different job types of the same user are aggregated to guarantee a fair distribution. In the following it is assumed that the system has knowledge of about � nodes and � jobs to be processed. The jobs belong to � individual types or users, respectively. Let �, �, � � � and ��. Furthermore • ��, ��1, … , �� denotes job �, • ��, ��1, … , �� denotes job type � (or user �, if ��), • ��: �1, … , �� 1, … , �� with �� =� denotes the type index of job � and • �� denotes the (exponentially weighted) average runtime of the finished jobs belonging to type ��, measured in wall clock minutes. Balanced Distribution Let �� be the number of jobs of type �� , which are running when a node asks for a job. The scheduler selects the job � which is minimizing �� . In other words, a node always gets a job of the type with the least jobs running. As a result, at any time every type/user gets the same number of nodes to work for it/him. This fairness only shifts when a user has not enough unprocessed jobs to keep the balance.
Page 2 and 3:
Lecture Notes in Computer Science 5
Page 4 and 5:
Volume Editors Christian Müller-Sc
Page 6 and 7:
General Chair Organization Christia
Page 8 and 9:
Organization IX Hartmut Schmeck Kar
Page 10 and 11:
Keynote Table of Contents HyVM - Hy
Page 12 and 13:
Table of Contents XIII JetBench:AnO
Page 14 and 15:
How to Enhance a Superscalar Proces
Page 16 and 17:
4 J. Mische et al. The Real-time Vi
Page 18 and 19:
6 J. Mische et al. 4.1 Instruction
Page 20 and 21:
8 J. Mische et al. Additionally the
Page 22 and 23:
10 J. Mische et al. % of unused pip
Page 24 and 25:
12 J. Mische et al. % of cycles spe
Page 26 and 27:
14 J. Mische et al. 16. Lickly, B.,
Page 28 and 29:
16 G. Aşılıoğlu, E.M. Kaya, and
Page 30 and 31:
Page 32 and 33:
Page 34 and 35:
Page 36 and 37:
Page 38 and 39: 26 T.B. Preußer, P. Reichel, and R
Page 50 and 51: 38 P. Bellasi, W. Fornaciari, and D
Page 62 and 63: 50 J. Zeppenfeld and A. Herkersdorf
Page 74 and 75: 62 B. Jakimovski, B. Meyer, and E.
Page 86 and 87: 74 M. Bonn and H. Schmeck Fig. 1. J
Page 90 and 91: 78 M. Bonn and H. Schmeck Uptime-ba
Page 92 and 93: 80 M. Bonn and H. Schmeck 2.4 Simul
Page 94 and 95: 82 M. Bonn and H. Schmeck done rate
Page 96 and 97: 84 M. Bonn and H. Schmeck Fig. 8. J
Page 98 and 99: 86 M. Bonn and H. Schmeck tells the
Page 100 and 101: 88 J.-P. Steghöfer et al. � �
Page 102 and 103: 90 J.-P. Steghöfer et al. mechanis
Page 104 and 105: 92 J.-P. Steghöfer et al. resource
Page 106 and 107: 94 J.-P. Steghöfer et al. Choose a
Page 108 and 109: 96 J.-P. Steghöfer et al. 1. Defin
Page 110 and 111: 98 J.-P. Steghöfer et al. and all
Page 112 and 113: 100 J.-P. Steghöfer et al. 19. Kim
Page 114 and 115: 102 K. Kloch et al. large-scale sys
Page 116 and 117: 104 K. Kloch et al. a�t� 1.0 0.
Page 118 and 119: 106 K. Kloch et al. constant. This
Page 120 and 121: 108 K. Kloch et al. Relative number
Page 122 and 123: 110 K. Kloch et al. (a) infection r
Page 124 and 125: 112 K. Kloch et al. (ii) Phase of a
Page 126 and 127: 114 P. Petoumenos et al. Studying t
Page 128 and 129: 116 P. Petoumenos et al. % of Misse
Page 130 and 131: 118 P. Petoumenos et al. IQ: n 4-in
Page 132 and 133: 120 P. Petoumenos et al. downsizing
Page 134 and 135: 122 P. Petoumenos et al. As long as
Page 136 and 137: 124 P. Petoumenos et al. comparable
Page 138 and 139:
Exploiting Inactive Rename Slots fo
Page 140 and 141:
128 M. Kayaalp et al. In a supersca
Page 142 and 143:
130 M. Kayaalp et al. INSTRUCTION 1
Page 144 and 145:
132 M. Kayaalp et al. time. Alterna
Page 146 and 147:
134 M. Kayaalp et al. Fig. 3. Numbe
Page 148 and 149:
136 M. Kayaalp et al. The results o
Page 150 and 151:
Efficient Transaction Nesting in Ha
Page 152 and 153:
140 Y. Liu et al. HTMs include TCC
Page 154 and 155:
142 Y. Liu et al. rollback T0 begin
Page 156 and 157:
144 Y. Liu et al. Processor core Pr
Page 158 and 159:
146 Y. Liu et al. 5.2 Results and A
Page 160 and 161:
148 Y. Liu et al. decreases, partia
Page 162 and 163:
Decentralized Energy-Management to
Page 164 and 165:
152 B. Becker et al. Furthermore, t
Page 166 and 167:
154 B. Becker et al. An optimizing
Page 168 and 169:
156 B. Becker et al. smart-home man
Page 170 and 171:
158 B. Becker et al. freedom like w
Page 172 and 173:
160 B. Becker et al. Power [W] 4500
Page 174 and 175:
EnergySaving Cluster Roll: Power Sa
Page 176 and 177:
164 M.F. Dolz et al. Themodulequeri
Page 178 and 179:
166 M.F. Dolz et al. This daemon al
Page 180 and 181:
168 M.F. Dolz et al. been submitted
Page 182 and 183:
170 M.F. Dolz et al. On the other h
Page 184 and 185:
172 M.F. Dolz et al. at the inactiv
Page 186 and 187:
Effect of the Degree of Neighborhoo
Page 188 and 189:
176 T. Abdullah et al. A zone based
Page 190 and 191:
178 T. Abdullah et al. A consumer/p
Page 192 and 193:
180 T. Abdullah et al. Messages 140
Page 194 and 195:
182 T. Abdullah et al. (except when
Page 196 and 197:
184 T. Abdullah et al. % Matchmakin
Page 198 and 199:
186 T. Abdullah et al. show that th
Page 200 and 201:
188 M. Schindewolf, D. Kramer, and
Page 202 and 203:
Page 204 and 205:
Page 206 and 207:
Page 208 and 209:
Page 210 and 211:
Page 212 and 213:
200 R. Plyaskin and A. Herkersdorf
Page 214 and 215:
Page 216 and 217:
Page 218 and 219:
Page 220 and 221:
Page 222 and 223:
Page 224 and 225:
212 M.Y. Qadri, D. Matichard, and K
Page 226 and 227:
Page 228 and 229:
Page 230 and 231:
Page 232 and 233:
Page 234 and 235:
A Tightly Coupled Accelerator Infra
Page 236 and 237:
224 F. Nowak and R. Buchty where A
Page 238 and 239:
226 F. Nowak and R. Buchty Fig. 3.
Page 240 and 241:
228 F. Nowak and R. Buchty Table 2.
Page 242 and 243:
230 F. Nowak and R. Buchty Table 4.
Page 244 and 245:
232 F. Nowak and R. Buchty 5.3 Comp
Page 246 and 247:
Optimizing Stencil Application on M
Page 248 and 249:
236 F. Xudong et al. compared to th
Page 250 and 251:
238 F. Xudong et al. stencil comput
Page 252 and 253:
240 F. Xudong et al. threads in the
Page 254 and 255:
242 F. Xudong et al. Speedup 14 12
Page 256 and 257:
244 F. Xudong et al. 5 Related Work
Page 258:
Abdullah, Tariq 174 Alima, Luc Onan
show all

Architecture of Computing Systems (Lecture Notes in Computer ...

Create successful ePaper yourself

Delete template?

Save as template?