From MIKE2 Methodology
--Sean.mcclowry 02:17, 27 February 2008 (EST) Much of this feedback is quite relevant across MIKE2.0 where the initial pass of mapping to the Usage Model was more conservatively applied. The key change to the approach would be to use the may be applied rating as opposed to the not required rating. Current activities should be re-assessed.
I've started to go through the ILM usage model - there's a lot of good information there! I've finished reading through Phase 1, and I had some comments I thought I'd share ...
Activity 1.2 Enterprise Information Management Awareness ... I disagree with the rating. It should be performed, but this is not a mandatory activity. The client stakeholders might have sufficient knowledge & understanding to make this an unnecessary step.
Activity 1.3 Overall Business Strategy for Information Development ... I disagree with the rating. This should definitely be performed at the beginning of a project, to establish the scope & the business objectives of a project. It allows for the project teams to hear different POVs without forming any preconceptions.
Activity 1.4 Organisational QuickScan for Information Development ... I disagree with the rating. This step should definitely be performed. This allows the project team to understand the client's current & desired state, which will influence the way the project is executed. Also, it's spelt "organiZational" ;)
--Satinder.parmar 17:24, 5 February 2008 (EST)The above changes would not affect inclusion in the project plan. Generally there would be aspects to a project that would drive the extent to which activities are performed and this would be at the discretion of the project team.
Activity 1.X ROI of Information Assets ... I disagree with the rating. This activity is typically not performed. It is difficult to capture the cost/ benefit parameters required for an ROI analysis at such an early stage. If anything, this calculation/ model can be used during the solution design phase, to compare solution effectiveness & desirability.
--Satinder.parmar 17:24, 5 February 2008 (EST)Note that the above activity is still under review and not formally a part of the task list.
I've been writing down some comments around the other phases too ... I'll upload them here when I finish them, and clean them up a bit.
--Sean.mcclowry 04:22, 6 February 2008 (EST) I agree and I have made the recommended changes. In terms of spelling, z to s, this is discussed here: National varieties of English
ILM Usage Model Phase 2
Here are my comments on Phase 2:
Activity 2.1 Strategic Requirements for BI Application Development ... I disagree with the rating. The requirements gathering process should be performed. It allows stakeholders concerns to be heard and garners support for the initiative.
--Sean.mcclowry 07:13, 23 February 2008 (EST) These are requirements apply to "front end" BI applications as opposed to the Strategic Requirements for Technology Backplane Development, which is required. This has note been changed. Does this make sense?
I'm not sure what you mean by "front end" BI applications. Regardless, a well designed system requires reporting, metrics, BI, etc. There may be existing BI systems (Crystal, BO) that need to be updated in order to poll information from the ILM application. Even if the BI mechanism is not changing, there maybe technological updates that facilitate BI integration. Hence, I still feel that this activity should be performed. - Akshay
- --Sean.mcclowry 01:11, 27 February 2008 (EST) By front end I mean apps like Business Intelligence applications generally as opposed to something specific within a BI suite. The front end is a reference to reporting and analytics as opposed to the database and ETL and data re-engineering components on the "back-end". To your point I will change this rating to may be performed.
Activity 2.6 Future-State Physical Architecture and Vendor Selection ... I disagree with this rating. It may be performed, as vendor selection is not always required - sometimes tools are preselected by existing implementations or contracts
--Sean.mcclowry 07:13, 23 February 2008 (EST) This activity can be a bit misleading as physical architecture and vendor selected are within a single activity. While vendor selection is not always required, physical architecture work is typically needed for large strategy projects. I have left the rating "as-is" but added a note to the activity - does this make sense?
Agreed - Akshay
ILM Usage Model Phase 3
Here are my comments on Phase 3:
Activity 3.9 Database Design ... I disagree with the rating. I think it should be "Typically Not Performed". Database design might be a sub-project within an ILM program; there can also be cases where a database would have to be re-engineered to support new ILM processes & techniques.
--Sean.mcclowry 07:13, 23 February 2008 (EST) Agreed - I have changed this activity as recommended above.
Activity 3.X Taxonomy Design ... I disgree with the rating. Depending on the maturity of the organisation and the nature of the project, defining the Taxonomy may be performed as an essential activity within an ILM initiative.
--Sean.mcclowry 07:13, 23 February 2008 (EST) Agreed - I have changed this activity as recommended above.
Activity 3.11 Data Profiling ... I disagree with the rating. Data profiling may be performed to assist in assessing data quality and identifying meta data.
--Sean.mcclowry 07:13, 23 February 2008 (EST) Agreed - I have changed this activity as recommended above.
Activity 3.12 Data Re-Engineering ... I disagree with the rating. Standardizing data and correcting data quality issues may be performed as part of an ILM initiative, (depending on the nature of the initiative).
--Sean.mcclowry 07:13, 23 February 2008 (EST) Agreed - I have changed this activity as recommended above.
Activity 3.15 Prototype the Solution Architecture ... I disagree with this rating. This step should be performed for any kind of system implementation. However, ILM projects are not necessarily system implementations ... even so, the "prototype" might be process related.
--Sean.mcclowry 07:13, 23 February 2008 (EST) Agreed - I have changed this activity as recommended above.
ILM Usage Model Phase 4
Here are my comments on Phase 4 ...
Activity 4.2 Business Intelligence Design ... I disagree with the rating. This activity may be performed, depending on the type of project.
--Sean.mcclowry 07:13, 23 February 2008 (EST) These are design applies to "front end" BI applications as opposed to the back-end of , which is required. This has note been changed. Does this make sense?
I'm not sure what your definition of "front end" vs. "back end" BI applications is. Could you please clarify this point? - Akshay
- --Sean.mcclowry 01:11, 27 February 2008 (EST) By front end I mean apps like Business Intelligence applications generally as opposed to something specific within a BI suite. The front end is a reference to reporting and analytics as opposed to the database and ETL and data re-engineering components on the "back-end". To your point I will change this rating to may be performed.
Activity 4.X Information Security Design ... I disagree with the rating. This activity should be performed, to review the security policies. However, it might not be a mandatory activity, given the scope of the project.
--Sean.mcclowry 07:13, 23 February 2008 (EST) I would think this would be critical in most cases for ILM when a new system solution is being built.
A new system solution does not always force a reorganization of information security policies. An ILM project might take the form of a new data archiving mechanism, or redeployment of storage architecture to implement tiering of resources. These don't require info sec design activities. I think the point I'm trying to make here is that given the broad scope of projects under the ILM umbrella, it makes sense to perform this activity, but it can be ruled irrelevant to the project at hand. - Akshay
--Sean.mcclowry 01:11, 27 February 2008 (EST) Agreed - a new system implementation may only make use of existing policies. On many projects where I have implemented new solutions this has been the case. I have added a note to the activity to help within this distinction.
Activity 4.3 Infrastructure Management Process Design ... 1. This activity is misnamed, given the description. According to the description, this is a Infrastructure Architecture Design activity, not a process design activity. 2. This activity is logically preceded by a Review, which is missing from this task list. That would give this task a "may be performed" rating.
--Sean.mcclowry 07:13, 23 February 2008 (EST) The tasks in this activity are procedural and architectural-oriented. The review is intentioned to be done earlier. Do you think possibly move some procedural stuff into a new activity and focus this on architecture? No changes yet ...
Having read through the description again, I'd like to retract point 1 above. However, I would like to raise a new point that the details of this activity are "mid-level" - they do not reference higher infrastructure management functions (e.g.: Configuration Management, Capacity Management, etc. - think IT Service Management functions), or lower functions (server management, network management). I think there needs to be some mention of these tie-ins. I like the fact that you've mentioned that this activity might leverage existing solutions.
I also feel that there needs to be a reference to the process review step that happens earlier - i.e.: which step in the usage model? - Akshay
- --Sean.mcclowry 01:11, 27 February 2008 (EST) This is an example on the infrastructural side of information management that may cross the boundary of what should go in MIKE2.0 (an IM workstream) vs. an infrastructure stream as per this model. As ILM and Data Center Management are now in-scope solutions for MIKE2.0, the importance of ITSM functions stand out in particular (although it would be of course important for other solutions. This is referenced in a minimal form in activities such as Software Development Lifecycle Preparation and Software Development Readiness. I think the best thing to do is to create an article explaining the overlaps for Information Management as a starting point - I will working on this ...
Activity 4.4. ETL Logical Design ... I disagree with the rating. This activity may be performed, depending on the type of project.
--Sean.mcclowry 07:13, 23 February 2008 (EST) Agreed - I have changed this activity as recommended above.
Activity 4.5 ETL Physical Design ... I disagree with the rating. This activity may be performed, depending on the type of project.
--Sean.mcclowry 07:13, 23 February 2008 (EST) Agreed - I have changed this activity as recommended above.
Activity 4.X User Interface Design ... I disagree with the rating. This activity may be performed, depending on the type of project.
--Sean.mcclowry 07:13, 23 February 2008 (EST) This related to developing a web interface. Are your thoughts that this would be for administration? No changes yet ...
The description of the task does not mention web-interface. Regardless of thin vs. fat client architecture, some thought must be put into making the system usable ... otherwise user adoption is going to be very low. The flip side is that the user interface might not require an update - instead the ILM project is to (for example) reorganize data asset categories. Hence, this activity may be performed. - Akshay
- --Sean.mcclowry 01:11, 27 February 2008 (EST) You are correct it does not make the distinction of thin vs. fat client - although many of our clients are web clients today this underlying tasks apply to either. I take your point re: usability - I have changed the rating and also added a note to the activity.
Activity 4.6 Services Oriented Architecture Design ... I disagree with the rating. This activity may be performed, depending on the type of project. This is especially true in cases where the use of ILM tools & techniques might impact/reorient existing SOA systems.
--Sean.mcclowry 07:13, 23 February 2008 (EST) Agreed - I have changed this activity as recommended above.
ILM Usage Model Phase 5
Here are my comments on Phase 5 ...
Activity 5.2 Technology Backplane Development ... I agree with the rating - but this should be broken up into smaller parts. This is the Activity where most of a project takes place and seems to be given disproportionately little weight. Would make "Implement Target Database" and "Develop Operationalised Data Re-Engineering Processes" as 5.X level Activities. This step should mention managing multiple environments - Development, Testing, Staging and Production."
--Sean.mcclowry 17:24, 18 February 2008 (EST) Yes, I agree this potentially causes the mis-perception that this is quite a small piece of work when it can be a very significant piece of the project. The downside of breaking this activity up is that it would made the phase even larger and details would be difficult to represent in a general sense. What I have done for now is added an explicit note to the activity regarding that warns the user.
Activity 5.3 BI Application Development ... I agree with this rating, but why is unit testing incorporated here instead of in a separate step as the other testing is?
--Sean.mcclowry 17:24, 18 February 2008 (EST) Unit testing is coupled with development in all activities. The rationale is that the developer performs unit testing and it should be done as part of their work activities for development and a set of test requirements that they have defined.
Activity 5.4 Functional Testing to Activity 5.7 Stress and Volume Testing ... there might be some merit to combining all these testing activities into one super-activity.
--Sean.mcclowry 17:24, 18 February 2008 (EST) The rationale behind keeping them separate was that the description of each is different, yet those differences tend to be the same across projects (and therefore it made sense to incorporate the detail into the Overall Implementation Guide. There is a chance this is a bit inconsistent to the point of development above.
Activity 5.8 User & Operations Training ... I agree with the rating, but would add flag for timing of testing: it can't be done too far in advance yet must be done prior to roll-out to collect feedback and update training or even re-train users.
--Sean.mcclowry 17:24, 18 February 2008 (EST) [[Functional Testing is actually quite early in the process and before much of the development is complete. It is the first step in the testing lifecycle after unit testing.
Activity 5.9 Production Deployment ... I agree with the rating, but would add back out plan as a deliverable
--Sean.mcclowry 17:50, 18 February 2008 (EST)Good point - I think it makes sense to have a backout plan, but do you think this should be earlier in the process? There could be significant business impacts of a solution doesn't deploy effectively ...
The reason why I highlighted backout plan at this step is because of a micro-view of Change Management - a backout plan is required by CM so that the changes made to the production environment can be rolled back in the event of change failure. At a higher level, there needs to be a backout plan for the business to ensure continuity ... this would have to be produced much earlier (maybe even during Design phase?) - Akshay
Activity 5.10 Evaluation and Launch ... I agree with the rating, but would put Final UAT in with Testing Activity
--Sean.mcclowry 17:50, 18 February 2008 (EST) This is actually covered through PVT, which provides a variation on UAT and typically uses a subset of the UAT test cases. I have added a note to this effect in the activity.
Activity 5.14 Continuous Improvement - Infrastructure ... I agree with the rating, but would add a flag that Continuous Improvement is a culture change that of which some Infrastructure staff may be suspicious
--Sean.mcclowry 17:50, 18 February 2008 (EST) Have added this comment to the activity
Activity 5.15 Continuous Improvement - Information Development Organisation ... I agree with the rating, but would add a flag that Continuous Improvement in Organization is a culture change that may threaten some staff as it might reduce their importance in the organization
--Sean.mcclowry 17:50, 18 February 2008 (EST) Have added this comment to the activity
Activity 5.16 Project Closeout ... I agree with the rating, but would add a flag that less mature organizations may not see the need for all closeout sub-activities
--Sean.mcclowry 17:50, 18 February 2008 (EST) Have added this comment to the activity
--Satinder.parmar 13:46, 12 February 2008 (EST)For comments that relate to the Activity itself, please note that these Activities and Tasks are applicable to a number of other offerings as well so that needs to be considered. If the comment applies to the ILM activity/task, please make the changes in the solution content article for ILM
Sean - Please give your thoughts on the rating changes and comments above for phases 2-5. Thanks.
BlogMarks
del.icio.us
digg
Furl
Simpy
smarking
