CC-IN2P3 Service Desk Project CC IN2P3 Service Desk Project - - PowerPoint PPT Presentation
CC-IN2P3 Service Desk Project CC IN2P3 Service Desk Project CC-IN2P3 Seminar, October 1, 2010 Holger Marten STEINBUCH CENTRE FOR COMPUTING STEINBUCH CENTRE FOR COMPUTING - SCC SCC KIT University of the State of Baden-Wuerttemberg and
STEINBUCH CENTRE FOR COMPUTING SCC
Holger Marten
STEINBUCH CENTRE FOR COMPUTING - SCC
KIT – University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association
www.kit.edu
2 Steinbuch Centre for Computing
October 1, 2010
/* ultimate solution */
if (apply_this_method) { (void) end_of_work; (void) end_of_talk; return 0; return 0; }
3 Steinbuch Centre for Computing
October 1, 2010
4 Steinbuch Centre for Computing
October 1, 2010
What is ITIL? What is the Service Desk? The CC-IN2P3 Service Desk Project
CC-IN2P3 specific:
Goals Requirements Technologies and Interfaces of the Service Desk Goals, Requirements, Technologies and Interfaces of the Service Desk
Ticket Analysis; Processes and Responsibilities of the Service Desk Ticket Analysis; Processes and Responsibilities of the Service Desk
Th S i D k d it CC (ITIL) di The Service Desk and its CC (ITIL) surrounding
5 Steinbuch Centre for Computing October 1, 2010
3 main goals of IT Service Management (ITSM)
Increase IT service quality (QM) q y (Q
)
Decrease long term cost of IT service delivery Arrange IT services according to current and future requirements of the enterprise (that’s us) and its customers (our user groups)
ITIL - The Information Technology Infrastructure Library
A Best Practice Recommendation for IT Service Management
First elements published in 1989 by CCTA of British Government World wide quasi-industrial standard
Approaches Service Management from the Lifecycle of a service D ib ITSM d d l f th i h dli Describes ITSM processes and recommends rules for their handling
Does neither instruct nor regulate rules or tools nor their implementation
Consequences Consequences
Steinbuch Centre for Computing October 1, 2010
(consisting of 27 processes, activities and functions) Service Strategy still correct?
ITIL V3
The “closed-loop problem”
Each phase based
p … but we are not building up on a green field!?
7 Steinbuch Centre for Computing October 1, 2010
The Service Desk (SD) is a functional unit of the ITIL Lifecycle Phase ( ) y ‘Service Operation’ The SD is also a Service that can be handled via the Lifecycle phases
Provide a single (central) point of contact for IT Users during daily work Resume (facilitate the restoration of) “normal operational service” to th IT U (C t ) ibl the IT Users (Customers) as soon as possible
Steinbuch Centre for Computing October 1, 2010
D i d i l t ti f th CC IN2P3 ifi S i D k Design and implementation of the CC-IN2P3 specific Service Desk
Provide the Service Desk with information and procedures that allow to reduce the work load of experts from support duties
(that they have to provide in parallel to planning operating and maintaining their services) (that they have to provide in parallel to planning, operating and maintaining their services)
Develop self-services by providing Users with first class tools and information... (such that they can find a solution of their problem by themselves) …to information... (such that they can find a solution of their problem by themselves) …to further reduce the work load of the Service Desk personnel and
Re-use grown CC-IN2P3 organizational structures, work methods, tools, etc. as far as possible
9 Steinbuch Centre for Computing October 1, 2010
Understand the ITIL Service Desk and its interfaces in general Understand the ITIL Service Desk and its interfaces in general Analyze the existing working methods of CC-IN2P3
Customer and User Support related communication, information, tools,… pp , , , Analyze ticket and workflows in xHelp
Categorize these according to ITIL
Id tif h t i S i D k l t d ll ITIL i t f Identify what is Service Desk related, as well as ITIL interfaces
Document the CC-IN2P3 specific design and implementation
Goals & work methods Goals & work methods Tools & technologies Functions, responsibilities and staffing Documents, processes and procedures Implementation Future developments and prospects currently draft v1.0
10 Steinbuch Centre for Computing October 1, 2010
What is ITIL? What is the Service Desk? The CC-IN2P3 Service Desk Project
CC-IN2P3 specific:
Goals Requirements Technologies and Interfaces of the Service Desk Goals, Requirements, Technologies and Interfaces of the Service Desk
Ticket Analysis; Processes and Responsibilities of the Service Desk Ticket Analysis; Processes and Responsibilities of the Service Desk
Th S i D k d it CC (ITIL) di The Service Desk and its CC (ITIL) surrounding
11 Steinbuch Centre for Computing October 1, 2010
Users have requests questions incident notifications
Users have requests, questions, incident notifications, … on the CC services and components during their daily business The Service Desk is the contact to deliver the solutions and provide answers answers
The Service Desk is the primary interface between Users and the CC The Service Desk is the primary interface between Users and the CC
SD ‘external view’ = main contact and (1st level) support SD ‘internal view’ = key role for User Information and Incident resolution
Daily User information must be channeled through the Service Desk The Service Desk must be well informed about current / recent Changes, Incidents, Problems, documentation,… g , , , , The Service Desk requires tools to do this The Service Desk requires processes (guidelines) to do this
12 Steinbuch Centre for Computing October 1, 2010
Resolution of
Documentation
Dynamic
Service Status
6 goals of the Service Desk
Resolution of external Requests and Incident Notifications Documentation Service Catalogue Dynamic Service Status Information Service Status Information Notifications Web interface to xHelp User Guide Charter of Usage Detailed docs, manuals Web interface to customized local monitoring (NAGIOS?) Mail, news, RSS feeds External reports to Customers
13 Steinbuch Centre for Computing October 1, 2010
Resolution of
Documentation
Dynamic
Service Status
Resolution of external Requests and Incident Notifications Documentation Service Catalogue Dynamic Service Status Information Service Status Information
User po
Notifications Web interface to xHelp User Guide Charter of Usage Detailed docs, manuals Web interface to customized local monitoring (NAGIOS?) Mail, news, RSS feeds
U
Service Desk process Service Catalog. User Access
Logical IT model Internal reports e.g. to CC
management see process guidelines xHelp engine Logical IT model CMS / DMS systems Local Monitoring engine (NAGIOS?) CCCom management,… CMS / DMS systems see Part II
14 Steinbuch Centre for Computing
October 1, 2010
User Portal = web services & servers + SPIP => port. util.
xHelp web interf. document links LMS web interface News / RSS xHelp engine CM / DM System . News, RSS systems Mail Systems LMS system CM / DM System . CM / DM S t
S Service Desk
CCCom engine
tools / applications
System .
S
how many?
Personnel
tools / applications
y
db / storage with solid backup
( ) 8h 5d
15 Steinbuch Centre for Computing October 1, 2010
24h x 7d (but certainly no on-call duty required) 8h x 5d
16 Steinbuch Centre for Computing October 1, 2010
17 Steinbuch Centre for Computing October 1, 2010
What is ITIL? What is the Service Desk? The CC-IN2P3 Service Desk Project
CC-IN2P3 specific:
Goals Requirements Technologies and Interfaces of the Service Desk Goals, Requirements, Technologies and Interfaces of the Service Desk
Ticket Analysis; Processes and Responsibilities of the Service Desk Ticket Analysis; Processes and Responsibilities of the Service Desk
Th S i D k d it CC (ITIL) di The Service Desk and its CC (ITIL) surrounding
18 Steinbuch Centre for Computing October 1, 2010
Steinbuch Centre for Computing October 1, 2010
Incident Management
An Incident is an unplanned interruption or reduction in the quality of an IT p p q y Service. Incident Management is the managing process to resume to normal operation as quickly as possible to minimize impact on (own and Customers) business. Translation: “Get this beast back online, no matter how!” Time scale: hours
Problem Management
A Problem is the cause of one or more Incidents. Problem Management comprises all activities to diagnose the underlying cause
incident again. Translation: “Uff, it’s back online. Now we can search for the error.” Time scale: days, weeks, …
20 Steinbuch Centre for Computing October 1, 2010
Password renewal after blocked login Notifications on Incidents already known by 2nd level (“yes, we have just changed… we are currently on maintenance … we’ve just detected a strange behavior…”) Incidents with User access changes, licenses, …
E l f ‘ i l ’ R t
Can be solved autonomously by the Service Desk on the basis of Service Desk Process Guidelines
Examples for ‘simple’ Requests
Request for explanation / documentation of services, commands,…
Problems
Users run into same Incident over several days or into (since weeks) well known issues again
Service Desk Process Guidelines
21 Steinbuch Centre for Computing October 1, 2010
C b l d b S i D k 1st l l t d
Can be solved by Service Desk = 1st level support under pre- conditions
(A) Information /documentation to Users is known (coordinated) by SD (see Part I of this talk) (B) Excellent internal info about Changes and detected Incidents
Esp.: Avoid system Changes without internal announcement! p y g Change Management
(C) Service Desk Process Guidelines available
SD identifies categories (e.g. from ticket analysis) and develops guidelines g ( g y ) p g together with experts
To be handed over to 2nd level support SD further follows up / asks back / coordinates if necessary
22 Steinbuch Centre for Computing October 1, 2010
p y
M t b h d d t Ch M t b t l f th
Must be handed over to Change Management but also further followed up by Service Desk
Someone must decide whether the Change is allowed and when it is done
Experts very often find and remove the error during Incident handling within hours
GOOD (if this is within the expected Service Level Agreement)! In other words: “Don’t forbid experts to search / remove the error during Incident resolution only because ITIL suggests so.” Incident resolution only because ITIL suggests so.
But: Hand all other Incidents, lasting more than 8 hours to Problem Management
23 Steinbuch Centre for Computing October 1, 2010
Handle User Request RfC ?
1
Inform User: “It’s a Change” Mark ticket internally as RfC
Handle Us Requests for C 4
yes Hand over to Change Management Request ‘simple’ ? Fix together with User
6
Solution possible?
ser Change Handle User Re
CLOSURE no yes yes p Fixed after p
‘simple’ equests Hand Use
CLOSURE no no Only
6
Hand over to 2nd li t
8 hours ?
dle ‘complex’ r Requests
yes no & workaround? yes no
6
CLOSURE 2nd line support & follow up Inform User: “It’s a larger issue”
Handle Problems 5
Mark ticket internally as Problem
3 2
Hand over to Problem Management
Weekly RfC and Problem Review 4 5
Check list of
Problem tickets Solution progresses ? Solved? Remove ticket from checklist
6
Escalate CLOSURE yes no
Steinbuch Centre for Computing October 1, 2010
Post ticket analysis = opportunity for self-improvement
Were Service Catalogue, descriptions, instructions, ... correct?
Did the User have the correct access rights? Were the Service Desk Process Guidelines available and correct? Request/Incident category for new Process Guidelines detected? W th tifi d I id t l k / i t d i t ll ? Was the notified Incident also known/communicated internally? Solution documented in FAQs & knowledge db? …
Definition and measurements of Target Metrics
What are our requirements? Once defined, measurements are straight forward.
How many tickets, <simple | complex> requests, incidents, problems? Overall turn around time until closure for these? Percentage of tickets solved by 1st level support? …
Training for Customers, CZARs, Users Customer feedback & next review Further ideas for self-service strategy?
25 Steinbuch Centre for Computing October 1, 2010
Further ideas for self service strategy?
What is ITIL? What is the Service Desk? The CC-IN2P3 Service Desk Project
CC-IN2P3 specific:
Goals Requirements Technologies and Interfaces of the Service Desk Goals, Requirements, Technologies and Interfaces of the Service Desk
Ticket Analysis; Processes and Responsibilities of the Service Desk Ticket Analysis; Processes and Responsibilities of the Service Desk
Th S i D k d it CC (ITIL) di The Service Desk and its CC (ITIL) surrounding
26 Steinbuch Centre for Computing October 1, 2010
1st level = SD Service Desk personnel Service Desk personnel 2nd level ‘Support Expérience’ Support Expérience CZARs Internal operations experts (net, storage…)
3 d 3 d 3rd level = 3rd party External experts (hardware vendors, software vendors e.g. ORACLE, SGE,…, ft d l id iddl ) software developers e.g. grid middleware)
27 Steinbuch Centre for Computing October 1, 2010
But please document activities in xHelp such that Service Desk is informed
Service Catalogue Management
Service Catalogue = most important document
g p
basis for common understanding of services, priorities, for structuring of all further documents & procedures,… Exists at >2 different places and specificities
1 t di t lid ti d d l t (Q lit M ?) 1 person to coordinate consolidation and development (Quality Manager?)
Change + Problem + Release and Deployment Management
SEPARATE from Service Desk (for hand-over and escalation) 1 person to coordinate and keep track of plans and announcements (IT Ops?)
Start with simple list of planning dates
Incident Management
Straight forward for external Incident Notifications (Service Desk) But handling and communication around internally detected Incidents? 1 person from IT Operations + Service Desk Manager together SEPARATE from Problem Management (as the latter is escalation path)
28 Steinbuch Centre for Computing October 1, 2010
Ticket
User Incident Notification
Nagios Alarm
Internal Event Notification
Alarm
User Support Operations Bridge
Did we recognize it as well? Which Services impacted? Who should take actions? Which Services impacted? Who should take actions? Whom else to inform? Requires long-term follow-up? Whom else to inform? Requires long-term follow-up?
Communication Collaboration Mutual support
29 Steinbuch Centre for Computing October 1, 2010
Mutual support
30 Steinbuch Centre for Computing October 1, 2010
an ITIL function and a visible Service
an ITIL function and a visible Service the daily point of contact and information for the CC Users internally a significant part of Request and Incident handling
it gets notice of all User issues
how many individual, non-registered mails & phone calls do we have?
it gets internal support from (almost) all departments
for tools, documents and information and communication
ticket post-analysis p y process guidelines clear definition of target metrics
31 Steinbuch Centre for Computing October 1, 2010
STEINBUCH CENTRE FOR COMPUTING SCC STEINBUCH CENTRE FOR COMPUTING SCC
holger.marten@cc.in2p3.fr
STEINBUCH CENTRE FOR COMPUTING - SCC STEINBUCH CENTRE FOR COMPUTING - SCC
KIT – University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association
www.kit.edu
KIT – University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association
www.kit.edu