Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 A Steering Portal for Condor/DAGMAN Naoya Maruyama on behalf of Akiko Iino Hidemoto Nakada, Satoshi Matsuoka Tokyo Institute of Technology.

Similar presentations


Presentation on theme: "1 A Steering Portal for Condor/DAGMAN Naoya Maruyama on behalf of Akiko Iino Hidemoto Nakada, Satoshi Matsuoka Tokyo Institute of Technology."— Presentation transcript:

1 1 A Steering Portal for Condor/DAGMAN Naoya Maruyama on behalf of Akiko Iino Hidemoto Nakada, Satoshi Matsuoka Tokyo Institute of Technology

2 2 Background Common Grid Usage Scenario  Zillions of Batch Jobs scheduled over combination of private/public resources within a VO Some Jobs require steering during workflow  “Human decision required” Most previous steering work focused on GUI- level interactivity  Real-time, interactive steering of the application itself  Does not meld well with batch jobs  Need significant application customizations

3 3 Objectives and Contributions Objectives  A Steering Portal for workflow (DAGMAN) jobs with easy descriptions, w/o application, Condor, or DAGMAN modifications Contributions  Portal to allow steering with simple additions to DAGMAN scripts  Confirmed low overhead with exemplar applications Quantitative assessment of user steps required

4 4 Outline Background Motivating example Required features of steering Steering example Overview and prototype implementation Evaluation Conclustion

5 5 Exemplar Application : Phylogenetic Tree Inference Infer phylogenetic relationships between different species from their genomic sequences [ Hasegawa&Shimodaira04] App Characteristics  Basically execute multiple parallel jobs in sequence => Workflow of batch jobs  But difficult to judge the termination condition of the application phases => Need human steering Common Ancestor

6 6 Narrow down on the candidate phylogenetic trees: Hard to automate=>batch jobs difficult Phylogenetic Tree Inference Breakdown Compute Posterior Probability “MrBayes” Compute likelihood value “PAML” Test “CONSEL”

7 7 List of Applications in the WF Job Descriptio n InputOutput Compute Time Required MrBayes Compute Posterior Probability Initial Topology List of Topologie s ~2 weeks on 24 high-end CPUs PAML Compute likelihood value List of Topologies Likelihood Values ~10 days on 26 high-end CPUs CONSELTest List of Topologies & Likelihood Values Probability Values 1~2 hours on 1 CPU

8 8 The Actual Workflow 1. Exec. MrBayes 2. Termination Judgement 3. Manutal input of new parameters 4. Post-Process MrBayes 5. Execute PAML 6. Execute CONSEL 55555 6 11111 2 Need Steering 3 4

9 9 MrBayes Example and Problems As a standalone app, requests interactive input  Up to a user to judge computational convergence But lacks info display to allow good judgment  Not on this screen! 1. User needs to periodically poll his screen and make interactive input 2. Also look at output files from 1000 jobs!

10 10 MrBayes Examples and Problems (2) Visualize ・ Decide on next parameter Problems: 3. Manual conversion to graphical display 4. Changing appropriate parameters Output file ・ Decide on Convergence

11 11 Outline Background Motivating example Required features of steering Steering example Overview and prototype implementation Evaluation Conclustion

12 12 Steering portal features for batch workflows with interactivity elements Pausing/resuming computation  Progress computation as much as possible until user input is absolutely needed  Resume immediately after input Allow flexible parameter modifications  Various ways to specify parameters for output and input  Various ways to notify users – interactive screen, email, etc.  Various ways of parameter observations – various portal functions  Various ways to modify parameters Even switching back and forth between your terminal and from a cell phone 10,000 miles away!

13 13 Outline Background Motivating example Required features of steering Steering example Overview and prototype implementation Evaluation Conclustion

14 14 Example: (1) Job submission Standard Condor/DAGMAN job submission  But includes steering functions in job description

15 15 Example (2): User Notification Various notification methods, incl. email Displays Portal URL in the message Works on various devices incl. cell phones

16 16 Example (3): Steering Portal Parameter Input Visualize current status Continuing of Workflow Portal generating steering web pages dynamically depending on workflow context

17 17 Outline Background Motivating example Required features of steering Steering example Overview and prototype implementation Evaluation Conclustion

18 18 Condor Pool Individual job submissions Workflow and Steering description DAGMAN/ Condor Steering–input Steering Portal User Notification Web page generation and Job control Overview of our Steering Portal submission Retry Function POST Scripting Features Steering– notification Steering–display

19 19 Overview of Steering Portal (2) The user defines several steering components for the steering portal, defining in a script below: A) A set of applications in the workflow B) CondorDAGMan+Steering workflow description A) Translator for converting output to input to continue workflow B) Visualization program to display application output on steering web page C) Application input/output specifications D) Parameters that require steering The Steering portal does:  Read the above script  Automatically generate steering web page  Interact with DAGMAN to notify users (email, etc.) and take input from the web portal

20 20 Prototype Implementation Coordination between DAGMAN and Steering Portal  Use DAGMan POST Scripting function to invoke the steering portal  Use DAGMan Retry function to resume workflow execution Prototype Implementation of the Steering Portal  Interpretation of the steering descriptions embedded in DAGMAN workflow  Appropriate and multiple notifications and steering interfaces available Notification and interfaces currently selected according to script  Automated selection for the future Mail and messaging notification function with embedded services CGI web page generation onto the portal sever using ssh Steering from anywhere, anytime (incl. cell phones and PDAs

21 21 Outline Background Motivating example Required features of steering Steering example Overview and prototype implementation Evaluation Conclustion

22 22 Evaluation Apply to sample applications (simple pi calculation and more complex phylogenetic tree example) Evaluate the necessary “work steps” Items of Evaluation A) Modification to the application program itself B) CondorDAGMan workflow description C) Translator for converting input to output to continue workflow D) Visualization program to display application output on steering web page E) Application input/output specifications F) Parameters that require steering G) Modifications to the Condor Job submit file

23 23 Sample Pi Program Eval. Item A No mod to the original program E Input: 4 inputs from stdin Output: 3 number columns F2 inputs out of the 4 stdin Eval. Item# Files # Lines in Total B24 C00 D13 G16

24 24 Phylogenetic Tree Program Eval. Item A No mod to the original program E Input: 1 setup file, 1 data file Output: 2 files F1 parameter value Eval. Item# Files # Lines in Total B36 C140 D116 G20 (1) 180 (1) 20 9-line files, only 1 line differs amongst them

25 25 Conclusion and Future Work Conclusion  Proposed a Steering Portal that allows interactive steering of batch scheduled jobs in Condor/DAGMAN  Created prototypes with flexible notification and visualization/steering features  Applied to sample apps including Pi and Phylogenetic trees Future work  Support and automatically select various interfaces  Apply to other application, esp. with larger workflows and more complex interactions  Apply to other workflow engines

26 26 Contact info  Satoshi Matsuoka, matsu@is.titech.ac.jp, Tokyo Institute of Technologymatsu@is.titech.ac.jp


Download ppt "1 A Steering Portal for Condor/DAGMAN Naoya Maruyama on behalf of Akiko Iino Hidemoto Nakada, Satoshi Matsuoka Tokyo Institute of Technology."

Similar presentations


Ads by Google