1. www.kit.edu KIT – The cooperation of Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH) ITIL and Grid services at GridKa CHEP 2009, 21 - 27 March, Prague Tobias König, Dr. Holger Marten Steinbuch Centre for Computing (SCC)

2. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) Main aim of GridKa is offering sustainable Grid services The availability and reliability of IT Services directly affects  The users‘ satisfaction  The reputation of the Computing Centre SCC and also not to forget the economical aspects Thus it is important to implement processes and tools that increase the availability and reliability of the IT services. The Information Technology Infrastructure Library (ITIL) is a process-orientated framework for the management of IT processes | Tobias König | CHEP 2009 | 21 - 27 March Prague2 GridKa: The German Tier-1 centre hosted by the SCC Steinbuch Centre for Computing: The Computing Centre of the KIT Introduction Karlsruhe Institute of Technology: The cooperation of the Forschungszentrum Karlsruhe GmbH (FZK) and the University of Karlsruhe (TH)

3. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) General information The SCC puts emphasis on ITIL v2 Service Support processes  Configuration Management  Incident Management and Service Desk  Problem Management  Change Management  Release Management Only these ITIL processes are currently implemented at the SCC. These processes are the most relevant and most important processes of ITIL for the SCC The other ITIL Service Delivery processes like  Service Level Management  Availability Management  Capacity Management  Continuity Management  Financial Management are not ITIL standardized and implemented at the SCC at the moment and in consequence they are not part of this talk | Tobias König | CHEP 2009 | 21 - 27 March Prague3

4. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) | Tobias König | CHEP 2009 | 21 - 27 March Prague4 Structure of this talk Incident Management Problem Management Change Management Configuration Management ITIL Processes at GridKa Respon- sibility (internal roles and groups) Internal view External view

5. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) | Tobias König | CHEP 2009 | 21 - 27 March Prague5 External/Internal Configuration Management Incident Management Problem Management Change Management Configuration Management Respon- sibility (internal roles and groups) Internal view External view ITIL Processes at GridKa

6. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) External/Internal Configuration Management The Configuration Management Database (CMDB) is the basis of all ITIL processes Following items are stored within the CMDB:  All the Configuration Items CIs (HW, SW, Racks)  The CIs are related to the service components  The IT services themselves are compositions of the service components.  Etc. | Tobias König | CHEP 2009 | 21 - 27 March Prague6 Configuration Management Respon- sibility (internal roles and groups) Internal view External view Service components Services CIs

7. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) | Tobias König | CHEP 2009 | 21 - 27 March Prague7 Incident Management Problem Management Change Management Configuration Management ITIL Processes at GridKa Respon- sibility (internal roles and groups) Internal view External view

8. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) Incident Management Prime service hours: Tickets are directly assigned from the 1st level support to the 2nd level support, the technical experts. The experts start immediately to solve the incident On-call service hours: An On-Call-Engineer (OCE) starts immediately to solve the incident Documentation of solution | Tobias König | CHEP 2009 | 21 - 27 March Prague8 Incident Management Problem Management Respon- sibility (internal roles and groups) Internal view External view Help Desk 1st Level Support Experts 2nd Level Support 5*12 Experts /7*24 On-Call Engineers

9. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) | Tobias König | CHEP 2009 | 21 - 27 March Prague9 Problem Management Incident Management Problem Management Change Management Configuration Management ITIL Processes at GridKa Respon- sibility (internal roles and groups) Internal view External view

10. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) Problem Management Prime service hours: Meeting of the task force (OCEs and Experts) On-call service hours: The OCE can call other experts the next morning or another OCE during the whole night Problem Management is the detection of the under- lying causes of an incident and their subsequent resolution and prevention. | Tobias König | CHEP 2009 | 21 - 27 March Prague10 Incident Management Problem Management Respon- sibility (internal roles and groups) Internal view External view Task Force: Problematic tickets, weekly reports, Availability, 7*24 Operations hand over Experts 2nd Level Support 7*12 Experts /7*24 On-Call Engineers

11. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) | Tobias König | CHEP 2009 | 21 - 27 March Prague11 External/Internal Monitoring Tools to support the Incident and Problem Management Incident Management Problem Management Change Management Configuration Management Respon- sibility (internal roles and groups) Internal view External view ITIL Processes at GridKa

12. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) External/Internal Monitoring Tools to support the Incident and Problem Management External: A incident or problem can be recognized from outside by the user or the external monitoring system. These alarms automatically create a ticket Internal: An incident or problem can be recognized from GridKa’s Nagios system Planned workflow between both systems | Tobias König | CHEP 2009 | 21 - 27 March Prague12 Incident Management Problem Management Respon- sibility (internal roles and groups) Internal view External view External Monitoring / SAM tests Internal Monitoring / Nagios, Ganglia External ticket system / GGUS Internal ticket system / Remedy 7*24 Knowledgebase / Wiki and Remedy

13. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) | Tobias König | CHEP 2009 | 21 - 27 March Prague13 Change Management Incident Management Problem Management Change Management Configuration Management ITIL Processes at GridKa Respon- sibility (internal roles and groups) Internal view External view

14. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) Change Management For changes of the IT infrastructure a Request for Change (RfC) is created in the CMDB There are 3 kinds of RfCs:  Emergency RfC  Standard RfC  Planned RfC GridKa’s downtimes are announced externally via EGEE Broadcasts and via the Change Calendar at http://www.gridka.de/monitoring | Tobias König | CHEP 2009 | 21 - 27 March Prague14 Change Management Respon- sibility (internal roles and groups) Internal view External view Change calendar for Maintenances Planned RfC Standard RfC Emergency RfC CMDB (RfCs)

15. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) | Tobias König | CHEP 2009 | 21 - 27 March Prague15 Special management roles Incident Management Problem Management Change Management Configuration Management ITIL Processes at GridKa Respon- sibility (internal roles and groups) Internal view External view

16. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) Special management roles The Configuration Manager  Responsible that all CIs are stored in the CMDB and that the DBs are up to date The Change Manager  Responsible for the change process and the formal correctness of the RfCs  Organizes the Change Advisory Board (CAB) Special role: Contact person to the IT Service management department | Tobias König | CHEP 2009 | 21 - 27 March Prague16 Change Management Respon- sibility (internal roles and groups) Internal view External view Change Manager CAB Change Advisory Board Configuration Management Configuration Manager GridKa Contact person to Service Management Collecting data for CMDB Supporting experts in filing RfCs

17. www.kit.edu KIT – The cooperation of Forschungszentrum Karlsruhe GmbH und Universität Karlsruhe (TH) Thank you for your attention! Steinbuch Centre for Computing (SCC)

18. KIT – The cooperation of Forschungszentrum Karlsruhe GmbH and Universität Karlsruhe (TH) | Tobias König | CHEP 2009 | 21 - 27 March Prague18 Discussion Incident Management Problem Management Change Management Configuration Management ITIL Processes at GridKa Respon- sibility (internal roles and groups) Internal view External view Change Manager Configuration Manager Task Force: Problematic tickets, weekly reports, Availability, 7*24 Operations hand over Help Desk 1st Level Support Experts 2nd Level Support 5*12 Experts /7*24 On-Call Engineers CAB Change Advisory Board GridKa Contact person to Service Management Collecting data for CMDB organizing Task Force and 7*24 operations Supporting experts in filing RfCs Change calendar for Maintenances Planned RfC Standard RfC Emergency RfC Service components External Monitoring / SAM tests Internal Monitoring / Nagios, Ganglia CMDB (CIs: Services, Service components, HW, SW, RfCs) SLA Services External ticket system / GGUS 7*24 Knowledgebase / Wiki and Remedy Internal ticket system / Remedy Workflow between Ticket Systems