Oren’s Weblog

Discussion of a SWOT analysis of help desk tools.

Our help desk tools are not well enough integrated, and terminology used in ITIL tends to be off-putting to many.

Knowledgebases are being used by many – mostly internally constructed and written. There’s at least one school that is using library school interns to work on the taxonomy of the knowledgebase.

There’s discussion of the relative merits of knowledge bases and wikis for support. The folks at Wisconsin note that their use of kb.wisc.edu, which is wiki-like for support of new technologies (Vista, Office 2007, IE7) has resulted in contributions of very high quality. Partner sites are motivated to be involved because they can take advantage of other people’s knowledge.

Chat – interactions tend to be longer, and staffing for it can be difficult. Industry says that agents can handle 3-4 chat sessions at a time if they’re dedicated to chat.

Discussion of customer satisfaction surveys – Greg objects to the use of the word “customer” in this context because the customer is the person who hands us resources to provide services, and when we survey people who use the services we aren’t talking to those people. It’s pointed out that ITIL defines “customer” as someone that funds the service. Bill Clebsch notes that at Stanford 2/3 – 3/4 of their business is fee-for-service, so the users are the customers.

Panel – demonstration of Duke using chat support by Debbie DiYula. Live chat link is on every web page within the OIT site. Getting lots of hits on chat – almost half have been telecom related (which is handled by a separate help desk). They now have configured chat to allow people to select which help desk they want to chat with. They have to manually input data from chat into the ticketing system.

There’s discussion about 24×7 support. Princeton is currently open 24×5 and moving to 24×7, but they don’t have a production or NOC team available. Bill Clebsch says that in Stanford’s survey of services and where people want them to invest, increased help desk hours come very far down the list. Carrie wonders if rural campuses have more desire for extended hours. Greg says that the only 24 hour operations on campus are the police and the hospital, so who are the people who need 24 hour support? 24×7 schools responding were Indiana, Minnesota, and Virginia Tech. Brad notes that they’ve been aggregating tasks to the places where the staff are already working the off shifts, like operations and the 24 hour student labs. One instutution says that their new network monitoring tools allowed them to move away from 24 hour support in the call center, as they didn’t have to rely as much on users being the “canary in the coal mine” for noticing problems.

Technorati Tags: CSG-Spring-2007, it-support, help-desks

We had a nice dinner cruise last night, leaving from the Chelsea piers and heading down around the lower tip of Manhattan and into the East River, under the Brooklyn, Manhattan, and Williamsburg bridges on a beautiful spring evening. The entertainment on board the Spirit of New Jersey was completely over the top in its enthusiasm and lack of quality, but I spent almost the entire cruise outside on deck enjoying the views and chatting with colleagues.

After the cruise Bob Morgan, Tony Chang, and I went to Birdland to catch Geri Allen’s 11 pm set with her quintet. It was really a treat to get to hear the almost eighty-year-old Jimmy Cobb on drums along with the young Rashaan Carter on bass, along with both Javon Jackson and Antoine Roney on tenors. I was particularly taken with her take on the ballads Lush Life (which opened the set) and Coltrane’s Naima, which had really nice bass playing from Carter.

Boy, you sure don’t hear 11 pm sets like this in clubs in Seattle on a Wednesday night. But it sure made for a hazy early morning today.

Technorati Tags: CSG-Spring-2007, jazz, music

We had a nice dinner cruise last night, leaving from the Chelsea piers and heading down around the lower tip of Manhattan and into the East River, under the Brooklyn, Manhattan, and Williamsburg bridges on a beautiful spring evening. The entertainment on board the Spirit of New Jersey was completely over the top in its enthusiasm and lack of quality, but I spent almost the entire cruise outside on deck enjoying the views and chatting with colleagues.

After the cruise Bob Morgan, Tony Chang, and I went to Birdland to catch Geri Allen’s 11 pm set with her quintet. It was really a treat to get to hear the almost eighty-year-old Jimmy Cobb on drums along with the young Rashaan Carter on bass, along with both Javon Jackson and Antoine Roney on tenors. I was particularly taken with her take on the ballads Lush Life (which opened the set) and Coltrane’s Naima, which had really nice bass playing from Carter.

Boy, you sure don’t hear 11 pm sets like this in clubs in Seattle on a Wednesday night. But it sure made for a hazy early morning today.

Technorati Tags: CSG-Spring-2007, jazz, music

Susan Grajek from Yale is kicking off the morning by presenting the results of a survey of CSG members about the help desk.

The survey results are on the web at:

http://www.stonesoup.org/Meeting.next/help.pres/grajek2.htm

Topics covered included the breadth of support, relationships with other support groups (both within and outside of central IT), service management, tools and best practices, and metrics. Did not ask about economics or costs of help desk operations.

Who is supported? All support faculty and staff, almost all support grad, undergrad, and postdoc students, far fewer support alumni, visitors, and clinicians. Students make up the preponderance of people supported internally.

Average help desk is open 14 hours/day on weekdays. Four are open 24 hours a day. A third are primarily staffed by students, which make up 64% of the workforce of those help desks. The help desks are open 7 hours/day on average on weekends. Half are closed on weekends. Students make up the entire weekend workforce.

Tools and best practices – All use some sort of ticketing or incident tracking system. Almost all use an automatic call distribution system. Most use some sort of knowledgebase or wiki. Two thirds use some sort of phone tree – seem to go one level deep, clients rebel at two levels deep and more than five choices. People are using data center and monitoring tools.

Almost a third are using remote desktop connections to provide in-depth support. Calls for a deeper knowledge level on the part of technicians answering calls.

Joel asks whether use of remote desktop for support raises security concerns. Greg reports that it does raise concerns, but they use it a lot.

Remedy is the most popular ticketing software.

Breadth of support – asked about 21 topics of support, could choose from 7 levels for each.

Email, network, web browsers, OS, accounts, and security prevention are the most heavily supported areas. High-performance computing, media, and library apps are the least deeply supported areas.

Some schools are providing help desk support for functional areas – the most common being use of course management tools, then financials, HR, and procurement.

Help desks are making handoffs to lots of other groups. Everyone are handing off to information security staff, and almost all to application and web developers and systems staff.

Almost half pass tickets to functional help desks like HR, Procurement, Library, health system, and classroom support.

two thirds provide access to tools for other IT support providers (knowledgebase, ticketing, systems status) and offer direct contact for these folks to back-room technologists.

Service management -

59% have different SLAs for different customer groups.

It’s noted in the back-channel notes that “we are concerned with improving routing of calls to the help desk which cannot be satisfied by front-line support; especially where people are essentially asking for special/new services — and what they really need is some really good consulting.” I think that’s a real issue – right now we tend to lose those requests for help.

Three quarters of schools have not evaluated outsourcing of help desk functions.

Pain points – responses include relationship with the ‘back-end’ of IT, the increasing complexity of support, upgrading tools, and staffing issues.

Technorati Tags: CSG, CSG-Winter-2007, help-desks, it-support, puppy

Bernie Gulachek from Minnesota is talking about how they’re demonstrating value of IT to the institution.

Why measure value? It’s not all about cost or efficiency, but it’s a balanced approach to show that the IT group is a strategic part of the institution, and can help the institution make strategic decisions. This is about making sure that IT has a seat at the table.

Look at definitions of value – bottom line (Gartner), educause documents.

IT Portfolio Management -

- understanding costs, and allocating costs to campus units, who don’t have a choice of what pieces they pay for. Once a year the units receive a bill that itemizes a rolled up version of what they’re paying (on a per-fte basis) for communications, productivity applications, business applications, and student, faculty, and staff support. That allows them to have a conversation of what the services cost (around $70 million/yr total, which works out to be $44.93/head/month). Portfolio services are framed in this way. That brings them to a conversation with the deans and IT directors in collegiate units – everyone is a stakeholder. If they have to pay for the services, then the value must be demonstrable. The administration is trying to get the institution to concentrate on the mission and to get to the right level of standardization by removing non-value-added duplication, and this allows that conversation to move forward.

Measuring the performance of the IT unit is done with the balanced scorecard construct. They’re breaking the conversation into four quadrants:

- service quality (measured by customer satisfaction);
- IT expenditure (productivity) (percentage of technology spend compared to total institutional spend – total IT spend, not just IT unit) – they can provide each collegiate and administrative unit’s budget spend on technology and what that percentage is, so that the deans can see the impact of what they choose to do locally. Works out to about $1682 per year per head (student, facult, and staff), of which about $600 is the central IT unit spend, the other $1000 is in the units. Good conversation about what the value is gained for that.
- Staff engagement – regular survey of staff satisfaction
- Improvement processes – developing tools for best practices for particular management disciplines and their adoption, plus understanding what the best practices for IT are and how they’re propagated to the units.

Impact of strategic initiatives / Projects

- what projects are out there that would be strategic to work on, given the impact to the institution. Working on institutional tools that normalize the language and measures of impact. Costs will always be realized in the IT world, but what are the functional benefits delivered to the functional unit. What they’re working on is a set of descriptive language that the President has been using to talk about these projects – frames up the elevator speech about specific projects. They walk through a work sheet form that highlight the impacts over the lifecycle of a project and where the costs are borne. Projections are over 3-5 year interval.

Example of an event ticketing system, where there are benefits realized in athletics and performing arts, and the costs are realized in IT.

Same process used for a shared data center project.

Benefit categories for checkoff include: improved productivity; reduced costs; enhanced revenues; improved service or product quality; engaged employees

IT Strategic Alignment to Support the Institution’s Goals

Strategy mapping is conceptual and fluffy, but it helps communicate.

The institution has set a goal of being one of the top three public research institutions in the next ten years and has set out four strategies to get there. IT has developed communication tools that demonstrate alignment with those goals. Simple maps that allow articulation of organization’s strategies and how they’re aligned with the institutions, and how the projects, measures, and operational goals are mapped to those strategies. This has led to a list of initiatives that they know are out there (whether defined or undefined) on a “pipeline” chart. This allows expression of all the requested initiatives. They’re working on an institutional process for deciding which initiatives to engage. The hope is to help decision makers understand priorities to weigh pet projects against the other data.

When projects are mapped to institutional strategies, they receive hugely positive response by presenting this in a very simple way from the President and other decision makers. Enables them to facilitate a conversation across different functional owners, encouraging them to work with each other. Allows them to see commonalities across projects, and maximize the use of scarce resources.

They’ve been asked to develop a collaborative informed decision and priority setting framework for large (more than $1 million) projects, similar to the process used for deciding building projects. The output from that would go into a full-blown analysis and RFI process.

Technorati Tags: business, CSG, CSG-Winter-2007, higher-ed, metrics

Bernie Gulachek from Minnesota is talking about how they’re demonstrating value of IT to the institution.

Why measure value? It’s not all about cost or efficiency, but it’s a balanced approach to show that the IT group is a strategic part of the institution, and can help the institution make strategic decisions. This is about making sure that IT has a seat at the table.

Look at definitions of value – bottom line (Gartner), educause documents.

IT Portfolio Management -

- understanding costs, and allocating costs to campus units, who don’t have a choice of what pieces they pay for. Once a year the units receive a bill that itemizes a rolled up version of what they’re paying (on a per-fte basis) for communications, productivity applications, business applications, and student, faculty, and staff support. That allows them to have a conversation of what the services cost (around $70 million/yr total, which works out to be $44.93/head/month). Portfolio services are framed in this way. That brings them to a conversation with the deans and IT directors in collegiate units – everyone is a stakeholder. If they have to pay for the services, then the value must be demonstrable. The administration is trying to get the institution to concentrate on the mission and to get to the right level of standardization by removing non-value-added duplication, and this allows that conversation to move forward.

Measuring the performance of the IT unit is done with the balanced scorecard construct. They’re breaking the conversation into four quadrants:

- service quality (measured by customer satisfaction);
- IT expenditure (productivity) (percentage of technology spend compared to total institutional spend – total IT spend, not just IT unit) – they can provide each collegiate and administrative unit’s budget spend on technology and what that percentage is, so that the deans can see the impact of what they choose to do locally. Works out to about $1682 per year per head (student, facult, and staff), of which about $600 is the central IT unit spend, the other $1000 is in the units. Good conversation about what the value is gained for that.
- Staff engagement – regular survey of staff satisfaction
- Improvement processes – developing tools for best practices for particular management disciplines and their adoption, plus understanding what the best practices for IT are and how they’re propagated to the units.

Impact of strategic initiatives / Projects

- what projects are out there that would be strategic to work on, given the impact to the institution. Working on institutional tools that normalize the language and measures of impact. Costs will always be realized in the IT world, but what are the functional benefits delivered to the functional unit. What they’re working on is a set of descriptive language that the President has been using to talk about these projects – frames up the elevator speech about specific projects. They walk through a work sheet form that highlight the impacts over the lifecycle of a project and where the costs are borne. Projections are over 3-5 year interval.

Example of an event ticketing system, where there are benefits realized in athletics and performing arts, and the costs are realized in IT.

Same process used for a shared data center project.

Benefit categories for checkoff include: improved productivity; reduced costs; enhanced revenues; improved service or product quality; engaged employees

IT Strategic Alignment to Support the Institution’s Goals

Strategy mapping is conceptual and fluffy, but it helps communicate.

The institution has set a goal of being one of the top three public research institutions in the next ten years and has set out four strategies to get there. IT has developed communication tools that demonstrate alignment with those goals. Simple maps that allow articulation of organization’s strategies and how they’re aligned with the institutions, and how the projects, measures, and operational goals are mapped to those strategies. This has led to a list of initiatives that they know are out there (whether defined or undefined) on a “pipeline” chart. This allows expression of all the requested initiatives. They’re working on an institutional process for deciding which initiatives to engage. The hope is to help decision makers understand priorities to weigh pet projects against the other data.

When projects are mapped to institutional strategies, they receive hugely positive response by presenting this in a very simple way from the President and other decision makers. Enables them to facilitate a conversation across different functional owners, encouraging them to work with each other. Allows them to see commonalities across projects, and maximize the use of scarce resources.

They’ve been asked to develop a collaborative informed decision and priority setting framework for large (more than $1 million) projects, similar to the process used for deciding building projects. The output from that would go into a full-blown analysis and RFI process.

Technorati Tags: business, CSG, CSG-Winter-2007, higher-ed, metrics

Dennis Maloney from Colorado introduces this workshop

What is the right approach?

Understanding the value of IT – Educause Quarterly November 2003
- Defining value, realizing value, structuring the value discussion, and metrics.

Brad Wheeler and Laurie Antolovic from Indiana are talking about Metrics, ABC, and the Value of IT to the institution. How big is the crew doing this activity? One financial analyst who spends half her time. The rest of the work is done by the people who are the custodians of the services.

Their annual process – Jul-Sep is assessment by the CIO and his cabinet, budget and project completeness, etc. Data collection takes place, using budget figures. The annual user survey is done then too, to provide user quality assessments. Laurie’s shop has really routinized all of this work. The survey is long and painful, but they get 40% response rate. They use the diagram of the annual planning processes in their budget hearings with the provosts at each of the campuses. It helps them frame the conversation, rather than having the conversation framed for them.

They can see lots of data on services, such as five year trends on service costs and unit costs for each service.

Pruning and reallocating – the expenditure review committee process trims budgets for reallocation to new priorities. Each manager has to give back a certain amount of money into a central pool for funding new proposals. Formal proposals request funding for new needs. Their latest was to take 5% from the entire base of the organization, across all the parts of the organization. They then can use that amount as being available for use to fund new proposals.

Difference between financial accounting and Activity Based Costing – in accounting money comes in to a unit, and then it is allocated for staff, equipment, etc. But that doesn’t tell you the cost of services, like what does it cost to provide an email account? or a support center call, or faculty consultation? This creates a healthy tension among service owners and providers. e.g. owner of peoplesoft hr system – “I need more attention from the sysadmins to tune my HR app” – DB manager: “I’ll charge 40% of sysadmin Joe’s time to the HR service owner” – you can throw a fit and go to the owner about that charge, or you can say “I always wanted that time, and now I’m going to use it.” Those conversations bring about a lot of behavioral change.

ABC came about as part of the quality movement.

Our first exercise is on establishing the cost of a service.

The example is the cost of support center activities.

Personnel
+ Direct Costs
+ Suppeliers (internal service providers
= fully-burdened cost

Divided by unit = service unit cost

Activities – answer help desk phone lines; assist walk-in customers; answer email help desk messages; maintain knowledgebase; provide extended consulting. Figured out for each person providing services

They already had a culture of distributed budget responsibility. That’s the level of detail from which they started. They spent a good chunk of time on defining services. Every service has an owner.

First they calculate the percentage of time each person spends on each activity. Has nothing to do with number of hours per week – just percentage of a person’s time overall. Then they derive the cost of each of those activities by dividing each person’s fully loaded full-time salary by the percentage of activities. They don’t do hour-tracking on a regular basis – just percentage estimates are good enough for what they’re doing.

Direct costs are figured using the percentage allocation by activities to allocate direct costs like hardware/software, training, tools, maintenance, etc. That approach doesn’t work in instances where there are large capital expenditures on equipment.

Internal services (internal systems administration, administrative support costs, etc) are allocated as Service Supplier Cost allocations into the fully burdened cost of goods.

Those total costs are then divided by the number of units served. It’s possible to have lots of arguments about what the units used for any given service should be. Use whatever unit best communicates the accountability and comparability of services.

For systems administration, the data center manager estimates the amount of effort that each system consumes, from example all of the operators’ time, etc.

Laurie runs training on ABC every year.

Costs are done on budget on a cash basis. Once you get over the initial cost, then everything is lifecycle funded. They’ve decided to not amortize capital expenditures for ABC to keep the complexity down. They’ve applied this to all services, regardless of the kind of money used to fund the service.

This doesn’t reflect the financial transactions – e.g. the people who run chat in the support center don’t pay money to the people who provide identity management, even though the id management people would allocate some of their activity to the support center for that activity in the ABC.

The cabinet goes through the analysis annually, and looks at anything that has less than 85% satisfaction rating for elimination (or for increased funding to improve).

Laurie’s slides from a 1997 CAUSE workshop on how to get started with ABC are online at http://www.stonesoup.org/Meeting.next/metrics.pres/CAUSE97.htm

Technorati Tags: business, CSG, CSG-Spring-2007, higher-ed, metrics

Dennis Maloney from Colorado introduces this workshop

What is the right approach?

Understanding the value of IT – Educause Quarterly November 2003
- Defining value, realizing value, structuring the value discussion, and metrics.

Brad Wheeler and Laurie Antolovic from Indiana are talking about Metrics, ABC, and the Value of IT to the institution. How big is the crew doing this activity? One financial analyst who spends half her time. The rest of the work is done by the people who are the custodians of the services.

Their annual process – Jul-Sep is assessment by the CIO and his cabinet, budget and project completeness, etc. Data collection takes place, using budget figures. The annual user survey is done then too, to provide user quality assessments. Laurie’s shop has really routinized all of this work. The survey is long and painful, but they get 40% response rate. They use the diagram of the annual planning processes in their budget hearings with the provosts at each of the campuses. It helps them frame the conversation, rather than having the conversation framed for them.

They can see lots of data on services, such as five year trends on service costs and unit costs for each service.

Pruning and reallocating – the expenditure review committee process trims budgets for reallocation to new priorities. Each manager has to give back a certain amount of money into a central pool for funding new proposals. Formal proposals request funding for new needs. Their latest was to take 5% from the entire base of the organization, across all the parts of the organization. They then can use that amount as being available for use to fund new proposals.

Difference between financial accounting and Activity Based Costing – in accounting money comes in to a unit, and then it is allocated for staff, equipment, etc. But that doesn’t tell you the cost of services, like what does it cost to provide an email account? or a support center call, or faculty consultation? This creates a healthy tension among service owners and providers. e.g. owner of peoplesoft hr system – “I need more attention from the sysadmins to tune my HR app” – DB manager: “I’ll charge 40% of sysadmin Joe’s time to the HR service owner” – you can throw a fit and go to the owner about that charge, or you can say “I always wanted that time, and now I’m going to use it.” Those conversations bring about a lot of behavioral change.

ABC came about as part of the quality movement.

Our first exercise is on establishing the cost of a service.

The example is the cost of support center activities.

Personnel
+ Direct Costs
+ Suppeliers (internal service providers
= fully-burdened cost

Divided by unit = service unit cost

Activities – answer help desk phone lines; assist walk-in customers; answer email help desk messages; maintain knowledgebase; provide extended consulting. Figured out for each person providing services

They already had a culture of distributed budget responsibility. That’s the level of detail from which they started. They spent a good chunk of time on defining services. Every service has an owner.

First they calculate the percentage of time each person spends on each activity. Has nothing to do with number of hours per week – just percentage of a person’s time overall. Then they derive the cost of each of those activities by dividing each person’s fully loaded full-time salary by the percentage of activities. They don’t do hour-tracking on a regular basis – just percentage estimates are good enough for what they’re doing.

Direct costs are figured using the percentage allocation by activities to allocate direct costs like hardware/software, training, tools, maintenance, etc. That approach doesn’t work in instances where there are large capital expenditures on equipment.

Internal services (internal systems administration, administrative support costs, etc) are allocated as Service Supplier Cost allocations into the fully burdened cost of goods.

Those total costs are then divided by the number of units served. It’s possible to have lots of arguments about what the units used for any given service should be. Use whatever unit best communicates the accountability and comparability of services.

For systems administration, the data center manager estimates the amount of effort that each system consumes, from example all of the operators’ time, etc.

Laurie runs training on ABC every year.

Costs are done on budget on a cash basis. Once you get over the initial cost, then everything is lifecycle funded. They’ve decided to not amortize capital expenditures for ABC to keep the complexity down. They’ve applied this to all services, regardless of the kind of money used to fund the service.

This doesn’t reflect the financial transactions – e.g. the people who run chat in the support center don’t pay money to the people who provide identity management, even though the id management people would allocate some of their activity to the support center for that activity in the ABC.

The cabinet goes through the analysis annually, and looks at anything that has less than 85% satisfaction rating for elimination (or for increased funding to improve).

Laurie’s slides from a 1997 CAUSE workshop on how to get started with ABC are online at http://www.stonesoup.org/Meeting.next/metrics.pres/CAUSE97.htm

Technorati Tags: business, CSG, CSG-Spring-2007, higher-ed, metrics

A panel with Paul Hill, Tony Chang, and Bruce Vincent.

Paul Hill kicks off the discussion with the obligatory “What is SOA” slide.

the first step is getting multiple groups to share a definition of what is meant by service.

MIT’s EAG guide has multiple definitions of service.

SOA is a set of design principles that decomposes common functionality into discrete services that can be used by a variety of systems.

It is not a technology.

Service Models – Burton talks about Infrastructure Service Model (authn, authz, auditing, logging, session management, data persistence, transactions) – enterprises try to expose these for their developers; and Business process management, that models the services on business functions instead of underlying technology, e.g. courses, purchase, add/drop, assessment, etc.

Bruce shows a diagram that Stanford has been working on to show services around collaboration tools.

MIT has done some work on categorization of services, but as they started to dive down into the technology to deliver web services, people have come forward with new services that they either need or can provide – e.g. the student services folks came forward and identified a need for geocoding (zip code to lat-long, and also internationally). They had never anticipated that need (and it’s interesting to note that the effort was catalyzed by the roadmap documentation effort).

Tony’s talking about how the UW is undertaking some pilot projects to build community around SOA in order to understand how we might engage as a university around those efforts.

Ken from NYU says that they’re working on an ITIL service catalog approach to listing the services that they offer, rather than taking on an SOA approach at present. Kitty notes that the admin systems group at Michigan is working with units to define data services around admin systems.

Lots of folks are working on building service catalogs – Chicago’s new one is at http://findit.uchicago.edu, Washington’s is at http://www.washington.edu/cac/planning/, Michigan’s old one is at http://www.itd.umich.edu/services/index.ph

The panel is now talking about obstacles to SOA – designing for reusability takes additional time and money – developers don’t know how to think about opening up their interfaces, don’t know how to plan for different toolsets than the ones they’re using. There’s a group at MIT called the IT SPARC group – they engage with projects that don’t have enough funding to take an enterprise approach to report the underfunding of projects that are too narrowly focused – they plea for one-time funding allocations to broaden scope. It’s been partially successful to date, but it’s early in the group’s existence.

Tony’s talking about the UW pilot SOA project, and how the approach is to try to understand how to build a framework that will encourage technology community around shared web services.

Bruce brings up the issue of maintaining backwards compatibility in services.

Paul points out that instrumentation of services is critical in terms of knowing who’s actually using services.

Brendan is talking about the fact that there are also data problems that may be exposed in the services – data definitions can change the meaning of data without notice to the users of services. This can be somewhat mitigated by strong data dictionary practices, but most universities haven’t been really strong in that area. As Tom Barton notes, this is not a new problem.

Paul says that many of the development units on campus haven’t yet made the transition to be able to consume web services, and are more comfortable dealing with traditional interface mechanisms like c libraries, jar files, etc.

Technorati Tags: CSG-Spring-2007, soa

A panel with Paul Hill, Tony Chang, and Bruce Vincent.

Paul Hill kicks off the discussion with the obligatory “What is SOA” slide.

the first step is getting multiple groups to share a definition of what is meant by service.

MIT’s EAG guide has multiple definitions of service.

SOA is a set of design principles that decomposes common functionality into discrete services that can be used by a variety of systems.

It is not a technology.

Service Models – Burton talks about Infrastructure Service Model (authn, authz, auditing, logging, session management, data persistence, transactions) – enterprises try to expose these for their developers; and Business process management, that models the services on business functions instead of underlying technology, e.g. courses, purchase, add/drop, assessment, etc.

Bruce shows a diagram that Stanford has been working on to show services around collaboration tools.

MIT has done some work on categorization of services, but as they started to dive down into the technology to deliver web services, people have come forward with new services that they either need or can provide – e.g. the student services folks came forward and identified a need for geocoding (zip code to lat-long, and also internationally). They had never anticipated that need (and it’s interesting to note that the effort was catalyzed by the roadmap documentation effort).

Tony’s talking about how the UW is undertaking some pilot projects to build community around SOA in order to understand how we might engage as a university around those efforts.

Ken from NYU says that they’re working on an ITIL service catalog approach to listing the services that they offer, rather than taking on an SOA approach at present. Kitty notes that the admin systems group at Michigan is working with units to define data services around admin systems.

Lots of folks are working on building service catalogs – Chicago’s new one is at http://findit.uchicago.edu, Washington’s is at http://www.washington.edu/cac/planning/, Michigan’s old one is at http://www.itd.umich.edu/services/index.ph

The panel is now talking about obstacles to SOA – designing for reusability takes additional time and money – developers don’t know how to think about opening up their interfaces, don’t know how to plan for different toolsets than the ones they’re using. There’s a group at MIT called the IT SPARC group – they engage with projects that don’t have enough funding to take an enterprise approach to report the underfunding of projects that are too narrowly focused – they plea for one-time funding allocations to broaden scope. It’s been partially successful to date, but it’s early in the group’s existence.

Tony’s talking about the UW pilot SOA project, and how the approach is to try to understand how to build a framework that will encourage technology community around shared web services.

Bruce brings up the issue of maintaining backwards compatibility in services.

Paul points out that instrumentation of services is critical in terms of knowing who’s actually using services.

Brendan is talking about the fact that there are also data problems that may be exposed in the services – data definitions can change the meaning of data without notice to the users of services. This can be somewhat mitigated by strong data dictionary practices, but most universities haven’t been really strong in that area. As Tom Barton notes, this is not a new problem.

Paul says that many of the development units on campus haven’t yet made the transition to be able to consume web services, and are more comfortable dealing with traditional interface mechanisms like c libraries, jar files, etc.

Technorati Tags: CSG-Spring-2007, soa