explore mongodb zhanggang 2013/05/24. detail analysis

22
Explore MongoDB zhanggang 2013/05/24

Upload: gervase-anthony

Post on 18-Dec-2015

229 views

Category:

Documents


2 download

TRANSCRIPT

Explore MongoDB

zhanggang

2013/05/24

Detail Analysis

Detail AnalysisROOT

◦Convenience, plot can ba changed dynamically

◦Not fiexible,data stored as root fileScript

◦Fiexible , get the data dynamically.◦A little complicated to gengrate a

good-look plot.

Detail AnalysisData retrieve

Detail AnalysisCPU efficiency

◦Script argument: field(F) value(V) python efficiency.py –F User –V 2 python efficiency.py -- field JobType --

value 10 python efficiency.py –F Site –V 5

Detail AnalysisCPU efficiency—per user

Detail AnalysisCPU efficiency—per user

Detail AnalysisCPU efficiency—per site

Detail AnalysisCPU efficiency—per site

Detail AnalysisCPU efficiency-JobType: MC

Simulation

Detail AnalysisJob major status

Detail AnalysisJob type: user(B),

reconstruction(G), MC(R)

Detail AnalysisDoubts

◦How to calculate Unnormalized CPUTime

Detail AnalysisExecTime VS DiskSpace

Detail AnalysisExecTime VS DiskSpace

Concurrency Test

Concurrency TestBackground scripts

◦sys_argv.py◦user.sh, site.sh, jobtype.sh◦run_client.sh

Call sys_argv.py

◦python sys_argv.py - F User - V 2 - N 50000

Concurrency Testrun *.sh

◦‘Field’.sh

run_client.sh

Concurrency Test

bg-jobs\times

1 2 3 4 5 avg(millis)

0 304 287 291 294 289 293

5 312 312 310 329 335 319.6

10 321 324 321 336 343 329

20 332 350 349 339 344 342.8

50 421 423 433 417 462 431.2

100 741 682 528 684 836 694.2

200 1523 1295 619 704 921 1012.4

Concurrency Test

Concurrency TestMongo status

Thanks