EGEE Asia Pacific Regional Operation Center Min-Hong Tsai ASGC - - PowerPoint PPT Presentation

EGEE-II INFSO-RI-031688

Enabling Grids for E-sciencE

www.eu-egee.org

EGEE Asia Pacific Regional Operation Center

Min-Hong Tsai

ASGC ISGC 2007 March 29, Taipei

http://www.eu-egee.org/

http://www.twgrid.org/aproc/

2

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Agenda

• APROC Introduction
• Status
• Joining EGEE

3

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

APROC Introduction I

• APROC Mission

– Provide deployment support facilitating Grid expansion – Maximize the availability of Grid services

• Supports EGEE sites in Asia Pacific since April 2005

– 20 production sites, 8 countries – 9 sites joined EGEE since last ISGC: recently HKU, KISTI – 3 sites in certification process

• Philippines: Advanced Science and Technology Institute
• Korea: KONKUK
• Mongolia: (MAS IPT) Mongolian Academy of Sciences

4

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

APROC Services

• Site Deployment Support

– Registration – Installation – Certification

• Operations Support

– Monitoring, troubleshooting – Problem tracking – Software updates and security coordination – Regional VO services - VOMS and LFC

• ASGCCA CA Service

– provide certificates for AP EGEE/LCG sites without domestic CA.

• EGEE Operations

– CIC-on-duty: EGEE global operations – Monitoring tool development: GStat and GGUS Search – TPM: Front line user support (Q4 2006) – OSCT: Incident Response duty (Dec 2006)

5

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

APROC Usage

• New Active VOs: Belle and TWGrid
• This year: 200 KSI2K Years
• Last year: 41 KSI2K Years

6

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

APROC Availability

• Daily snapshots of SAM results of

region

– Availability increased to 70-80% range from 60-70% a half year ago

• CT mostly replica management

failure

– Sensitive to Information System access/performance – Request that data management clients can failover to secondary BDII

• Network Issues

– Often the root cause of CT, JL and JS – Network congested site set up local top- level BDII

• Increase default update timeout and

breath time

20 40 60 80 100 2005-04 2005-06 2005-08 2005-10 2005-12 2006-02 2006-04 2006-06 2006-08 2006-10 2006-12 2007-02 avail avail reliab 0% 20% 40% 60% 80% 100% 2005-04 2005-07 2005-10 2006-01 2006-04 2006-07 2006-10 2007-01 SD CT JL JS ER OK

2.4 2.6 2.7 3.0

Remove Slow BDII JS from SSH upgrade LHC OPN Hardware Failure

7

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Monitoring and Notification

• Planned integration of Asset DB
• Nagios plugins developed
• CE
• LFC
• VOMS
• Storage
• IT services
• OS
• Notification via Email

– SMS transmission device currently being tested

8

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Nagios Regional Monitoring

• Tests run at faster frequency

– 5-10 minutes – Faster response to faults

• Add customized plugins

– Run low level tests for faster isolation of problems – Tests may not be available in global monitoring tools yet – Ability to run tests on the target host via NRPE

• Management Interface

– Acknowledgement – On demand execution of tests – Historical availability – Test dependencies

http://lists.grid.sinica.edu.tw/apwiki/Nagios_monitoring_-_APROC_sites http://lists.grid.sinica.edu.tw/apwiki/Nagios_Plugins_Description

9

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Plans

• Increase monitoring coverage

– Information System – Network performance monitoring

• available/achievable bandwidth
• Full mesh monitoring
• Improve troubleshooting tools

– http://lists.grid.sinica.edu.tw/apwiki/APROC/Troubleshooting_Guides – FAQ system – Service diagnostic scripts

• Integration of ticketing system with GGUS
• Training

– EGEE Induction at GridAsia 2007. June 5, 2007 Singapore.

10

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Joining EGEE Infrastructure

• Contact APROC
• If domestic CA is not available

– Register as a ASGCCA RA during ISGC

• Dedicated an administrator with Unix experience
• Allocate servers

– 5: UI, CE, WN, DPM, MON – 3: CE/WN, MON, DPM

• UI can be installed in user account
• Consider Virtual Machine for MON
• Study user guide and installation manual
• Send configuration file to APROC for review before deployment
• Complete registration and certification process

11

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Long Term Operations

• Establish domestic CA if none exists
• Increase availability and resource levels
• Establish domestic operations structure

– Operations procedures – Tools: monitoring and notification, ticketing system – User and administrator support

• Training for administrators and users
• Collaborate with APROC in Regional operations
• Q: Need for regional experimental Grid?

12

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Issues in AsiaPacific

• No regional projects to promote collaboration in EGEE
• Network bandwidth

– Low capacity: regional and last mile – Usage based billing

• Need for training

– Training for trainers – Application Training – E-Learning material

• However EGEE already provides

– M/W development and integration – Operations structure, coordination and support – Close to 200 user communities

13

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Summary

• APROC Provides EGEE operations support services to AsiaPacific
• EGEE sites in region has grown to 20 sites with utilization of 200

ksi2k years

• We have also improved availability but still is significant room for

improvement

• We look forward to more site joining EGEE in the region and eht

possibility for further collaboration

– Applications – Operations

• Feedback on what we can improve

14

Enabling Grids for E-sciencE

EGEE-II INFSO-RI-031688

Thanks You for Your Attention!

• Questions?

– roc@lists.grid.sinica.edu.tw – http://www.twgrid.org/aproc/

• Thanks to efforts from:

– T1/APROC Team

• Jason Shih

Dave Wei

• Felix Lee

Joanna Huang

• Aries Hong

Hung-Che Jen

• Jinny Chien

Shu-Ting Liao

• Yi-Ping Wu

Min Tsai