Author: Martin Kirk
Modification Date:
19/08/03 22:04
Martin Kirk
The Open Group
m.kirk@opengroup.org
Larry Wittorff
Boeing
larry.wittorff@boeing.com
Tom Studwell
IBM
studwell@us.ibm.com
Tom Bishop
Vieo
tbishop@vieo.com
Carl Bunje
Boeing
carl.f.bunje@boeing.com
Todd Gamble
MCI
todd.gamble@mci.com
Shinya Aihara
Fuji Research Institute Corporation
shinya_aihara@fuji-ric.co.jp
Neil Davies
Predictable Network Solutions
neil.davies@pnsol.com
Fred Hammond
Predictable Network Solutions
fred.hammond@pnsol.com
Jim Bouhana
Fujitsu Australia Ltd
james.bouhana@fujitsu.com.au
Mike Tsykin
Fujitsu Australia Ltd
mike.tsykin@fujitsu.com.au
Ron Higgin
BMC Solutions
ron_higgin@bmc.com
Jean Hammod
JPH Associates
jph@jhammond.com
Sally Long
The Open Group s.long@opengroup.org
Karl Schopmeyer
Inova
k.schopmeyer@opengroup.org
Barbara DeLibero
IBM
barbde@us.ibm.com
1. Introduction
The meeting started with a reprise of Tom Bishop's presentation from April
During the subsequent discussion. Jean noted that old applications won't disappear. There will be a need to front-end and encapsulate them. TomB observed that the new model is what is driving complexity up. Todd pointed out that old mgmt frameworks aren't going away either!
TomB stressed the point that there is no single cohesive stands body. TomS observed that the wants to, but isn't doing it all.
Jean mentioned that one of the plenary speakers in the main conference noted things tend to drive down to the lowest cost solution. Web services may not be the solution everywhere owing to XML weight.
Neil questioned the point about needing to manage to tighter tolerance, asking to what purpose? If the purpose is sufficiently rich, we can afford a broader set of options.
TomB explained that he had heard from many CIOs that hardware budgets are frozen. Their need to get utilization up. (From 10% to 20% would be a great improvement!) There is a great need to drive out complexity. There are a lot of groups working on related things. AQRM focuses on the bigger picture across all the environment.
Ron felt that there were a lot of different groups, centered around individual technologies. No one seems to be concentrating on getting them all to come together. Proprietary solutions tend to trade on their unique characteristics, and will be loathe to give them up. TomB stressed that putting in a right level solution enables rather than stifles innovation. Look at TCP/IP. Martin observed that the key is defining standards at the right level.
Ron stated that SNMP is universal for hardware, but we need something for applications.
TomS felt that the problem statement addressees business focus rather than individual small components. The customer needs for management should be factored in. We need other organizations to work on their management.
TomB observed that successful standards all seem to be based around consensus on a set of layers of abstraction. If we can look around the industry, pick all the standards to populate our profile, that would be wonderful. A lot of liaison will have to be with non-management groups, rather than groups defining infrastructure.
Carl felt that the key to success is to promote/evangelize.
The result of the discussion was some modifications to the group's charter, which now reads:
To establish and promote:
- a set of architecture principles,
- profiles of standards, and
- appropriate standards
for an adaptive approach to measuring and controlling the Quality of Experience and the Quality of Service for business functions (delivered by existing and new applications) across one or more business-responsive enterprises.
The objectives of the group will include:
- Recognition / acceptance /endorsement of an adaptive approach to managing distributed applications and supporting infrastructure.
- The model must include business, operations and resource perspectives.
- Identification of standards gaps and an action plan to close gaps.
- Industry ratification of architecture principles and profiles of standards for the above objectives,
- Ensure management interoperability, across management domains, and plug-and-play integration, within resource domains by funding and hosting interoperability testing.
Milestones for meeting these objectives will include:
- Standard measures of Application Quality of Service
- An Application Model-Based Management Architecture
- Measure/Affect instrumentation (standards framework)
An important aspect of the work of the group will be that it will work cooperatively with other groups and will make use of existing or emerging standards wherever possible.
2. Formation of Sub-Groups
The group then moved on to form a series of subgroups, as follows:
- Architecture
- Policy
- Application Instrumentation
- AQoS
- Industry Research
- Recruiting
The subgroups proceeded to work on their own and reported back towards the end of the meeting.
To support the activities of the subgroups, Martin reorganized the website and create additional mailing lists. Each sub-group has its own work area and mailing list.
3. Presentations
The group received 2 presentations. The first, titled "Towards Standards for Dynamic Resource Management" was from David Fleeman, Ohio University. The second, from Neil Davies, was titled "Understanding Applications Quality's Interaction with Resource Management". Both presentations are posted on the AQRM website.
4. Reports Back
Reports back from each of the subgroups were presented as follows:
Architecture - Karl Schopmeyer
Policy - Jean Hammond
Application Instrumentation - Larry Wittorff
AQoS - Neil Davies
Industry Research - Karl Schopmeyer
Recruiting - Jean Hammond
Presentations form each subgroup are posted on the AQRM website.
5. Future Actions and Organization
Now that we have established the subgroups, they will be supported by:
- subgroup mailing lists
- work areas for each subgroup
- closed part of website writeable/accessible to subgroup only
- Wider access available to all AQRM participants
public areas accessible to all
The current interim Steering Committee activity will be folded in to the main list for the time being. This will involve all participants in the steering/planning activities.Future Teleconference Schedule:
Date Topic Notes Aug 5 Steering Aug 12 Policy Aug 19 Steering / Industry Research Architecture? Aug 25 Possible Face-to-Face in Washington Aug 26 Recruiting Sep 2 Steering / AQoS Sep 9 Policy Sep 16 Steering / Architecture Sep 23 Application Instrumentation Sep 30 Steering / Industry research Oct 7 AQoS Oct 14 Steering Oct 22-23 Face-to-Face - Washington Conference