CHEP 2000

1
CHEP 2000

• Smart Resource Management Software
• in High Energy Physics
• Wolfgang Gentzsch and Lothar Lippert
• Gridware GmbH Inc.
• Padua, 9 February 2000

2
CHEP 2000 Resource Management with CODINE / GRD
Technical Requirements and Features

• what do we offer to help HEP Computing

Gridware - The Company

• Technology Leader in Resource Management

A special offer to the HEP community

• Our answer to falling hardware-prices

3
Technical Requirements and Features

• Array Jobs
• Advanced Queue Concept
• Policy Management
• Separation of Components
• Solutions for mixing interactive and batch
• Simplified system administration
• AFS Support
• CORBA Interface
• All classic Features
• Availability

4
Array Jobs
!/bin/sh ...
1 single Submit-Command for thousands of similar
jobs

• Example qsub -t 1-10001 jobscript.sh
• creates 1000 instances of a single job
• The whole array can be (also partly) manipulated
  (deleted, suspended, ...) with 1 command
• unlimited number of instances

5
Advanced Queue Concept
Emergency Room Concept
Grocery Store Concept
Job
Q1
Cluster
Job
Job
Dispatch
Q2

• The whole cluster can be adressed
• Soft requests are supported
• No empty queues while others are more than
  full
• each host can be treated with different
  policies
• users just request resources

• Cluster is split
• Queues may run empty
• users have to decide for a queue
• Job has to stay in line also if other
  resources are unused

Example qsub -q 10MQ jobscript.sh
Example qsub -l mem_free10M jobscript.sh
higher efficiency
6
Policy Management
Fairshare
Override System
20 Group1
Boosts temporarily project/job/group/department
30 Group2
50 Group3
Raise group
Share Utilization
Time
7
Separation of Components
Separation of Master and Scheduler

• Scalability
• high performance
• good response time
• faster job placement

8
Simplified system administration
Conifiguration changes without any pain

• No daemon restarts necessary
• Add machines on the fly
• Ability to install the entire cluster from one
  workstation
• No submit daemons or configuration needed for
  client
• Optimized architecture provides reliability

9
What else?
All classic Features

• accounting, monitoring, suspension, sensors ...

Interactive vs. Batch

• time windows
• automatic suspend
• migration, ...

Availability

• all leading unix platforms

CORBA Interface AFS Support
10
The company
GENIAS
Chord

• based in Germany
• European Union funded projects
• RD company

• located in California
• leader in sales of RMS

• Technology leader in Resource Management
• Goal make CODINE world standard in Resource
  Management

11
Our experience
EU funded research projects

• REMUS
• UNICORE...

Reseach Development

• DESY Zeuthen (long relationship)
• CASPUR (recently switched to CODINE)
• MPI (Max Planck Institutes)
• ...

Industry

• BMW
• SAAB
• SIEMENS
• ...

12
Contact Us
http//www.gridware.de mbox_at_gridware.de 49 (0)
9401 92 00 0 lothar.lippert_at_gridware.de