Http:// // NGS computation services: API's,

http://www.ngs.ac.ukhttp://www.grid-support.ac.uk

http://www.eu-egee.org/http://www.pparc.ac.uk/http://www.nesc.ac.uk/

NGS computation services: API's, concurrent and parallel

jobs Mike Mineter

mjm@nesc.ac.uk

3

Overview

3 distinct techniques will be in the practical:

• API’s to the low-level tools

• Using multiple processors– Concurrent processing - for many jobs– Parallel processing – to speed execution of one job

4

Overview

• API’s to the low-level tools

5

Job submission so far

GLOBUS, etc.

Command-line interfaces

Portal

Browser

User’s Interface to the grid

GLOBUS, etc.

6

Application-specific tools

GLOBUS, etc.

Command-line interfaces

Portal

Browser

Application Specific

and / or

Higher generic tools

Portal

Browser

User’s Interface to the grid

7

GLOBUS, etc.

Application Specific

and / or

Higher generic tools

Application-specific tools

Command-line interfaces

Portal

Browser

User’s Interface to the grid

API’s:

•Java

•C

•…

8

Available API’s

• “Community Grid” CoG http://www.cogkit.org/– Java, Python, Matlab– (very limited functionality on Windows – no GSI)

• C http://www.globus.org/developer/api-reference.html – Example in the practical

9

Using multiple processors

10

High performance computing

• If run-times are too long, you can – Optimise, re-code – Use HPC

• Parallel processing: multiple processors cooperate on one task – you run one executable; its distributed for you

• Concurrent processing: multiple processors, multiple tasks – multiple executables distributed and run by a script you write

• High-throughput computing: multiple processors, multiple tasks using spare CPU cycles - CPU years in elapsed days. E.g. sensitivity analyses in modelling. Condor, for example.

11

High performance computing

• HPC– Parallel processing: on NGS, you run one globus-job-

submit command – but need to code and build program so it is parallelised

– Concurrent processing: on NGS, multiple executables run from a script on the UI

– High-throughput computing: Many independent runs using spare CPU cycles - CPU years in days.

NGS core nodes open these routes to you – but you have to do a bit of work! (Grid is not magic!...)

Watch for addition of Condor pools on the NGS!!!!

12

Concurrent processing

UI

Internet

Head processors of clusters

Worker processors of clusters

Globus_job_submit

via PBS

Processes run independently

13

Concurrent processing

• An approach:– Script 1: to submit multiple jobs and gather their URIs

from globus-job-submit– Script 2: to test for completion and gather results

– In the practical, the jobs are short, so these scripts are integrated

14

Parallel Processing

• In early 1990’s parallel processing had some similarities to current state of grid computing – if on a smaller scale– National resources begun to be available– Standards were needed– Lot of hype– Early adopters in physics– Inadequate integration with the desktop

• And from the jungle of conflicting approaches emerged…

• MPI: Message Passing Interfacehttp://www-unix.mcs.anl.gov/mpi/mpich/

15

Parallel processing

UI

Internet

Head processors of clusters

Worker processors of clusters

Globus_job_submit

16

Parallel processing

UI

Internet

Head processors of clusters

Worker processors of clusters

MPI

17

Parallel processing

UI

Internet

Head processors of clusters

Worker processors of clusters

MPI

Processes communicate, synchronise(the simpler the better!)

18

MPI concepts• Each processor is assigned an ID [0,P-1] for P processors

– Used to send/recv messages and to determine processing to be done• Usually code so that processor 0 has coordinating role:

– Reading input files, Broadcasting data to all other processors– Receiving results

• Point-to-point Messages are used to exchange data– Either blocking (wait until both processors are ready to send/receive)– Or non-blocking: start the send/recv, continue, check for send/recv

completion• Collective functions

– Broadcast: sends data to all processors– Scatter: sends array elements to different processors– Gather: collects different array elements from different processors– Reduce: collects data and performs operation as dta are received

• Can group processors for specific purposes (e.g. functional parallelism)

19

MPI notes

• How does the task split into sub-tasks?– By functions that can run in parallel??!– By sending different subsets of data to different processes?

More usual ! Overheads of scatter and gather• Need to design and code carefully: be alert to

– sequential parts of your program (if half your runtime is sequential, speedup will never be more than 2)

– how load can be balanced (64 processes with 65 tasks will achieve no speedup over 33 processes)

– Deadlock! • MPI functions are usually invoked from C, Fortran

programs, but also Java• Several example patterns are given in the practical.

Many MPI tutorials are on the Web!

20

Summary

• Can build user applications and higher level tools on GT2 via APIs

• NGS core nodes support– concurrent processing (if you do a bit of scripting /

coding)– MPI for parallelism (if you parallelise your code)

• Remember – the CSAR, HPCx services!– NGS is batch oriented at present.

21

Practical

• http://homepages.nesc.ac.uk/~gcw/NGS/GRAM_II.html

• (That II is letters not numbers on the end!)