Notes
Outline
Management
W3C Face To Face
11/12/02
Management Task Force
Goal: Draft architecture to satisfy management requirements
TimeLine: 4-6 weeks
Deliverables:
Basic Management Data
Access to Management Data
List of Outstanding To Dos and Timelines
Approach to coordinate with other Standards Bodies
Participants and Progress
Participants
Currently interested parties:
IBM – Heather Kreger
Computer Associates – Igor Sedukhin
HP – Zulah Eckert, Yin Leng Husband
Thomson – Hao He
Talking Blocks – Mark Potts
WebMethods – Sandeep Kumar
Sonic – Colleen Evans
W3C WSAWG’s role
These works need a Web Services Architecture
To give them context
To give them guidance
To keep deliverables in consistent
To promote interoperability/cooperation
I asked for 6 weeks for the W3C to make progress
I have to report back to the DMTF (Tues) and OASIS (Wed) groups next week
Outside Forces
OASIS
Management Protocol TC: Access to management data using Web Services
Talked to OASIS MPTC: glad to accept and submit requirements
Groundswell to charter another: Management of Web Services
Groundswell has quieted, at least till W3C WS Arch MTF is done
DMTF
Was Kicking off new modeling effort: Modeling web services management.
Put modeling WS on hold till W3C WS Arch MTF is done
Manageable Components
Service Environment
For the Container
Hosted Service
For its contained Services
Service
Discovery Agency
TBD: Intermediary, Gateway, proxy…
TBD: Client Environment, Client
Progress
Set of conversations and consensus on some concepts that were not unique to management brought back to WSA
Provider detail
Requester detail
Discovery agency detail
Relationships between roles (UML)
Service Lifecycle
Request Lifecycle
Progress
For each component (no consensus on specific data)
Data (ID, Configuration)
Metrics
Operations
Events
Describe using Schema and WSDL
Access to manageability data
To Service Environment, Service, Discovery agency
WS defines access to information
Use WSDL to describe bindings
Discovery of manageable resources
Use WS discovery agencies to discover WSDL and manageability of services
Management pattern
Manageable Service
Vote?
Continue the MTF till the next F2F
Achieve consensus on:
Base operations, metrics, events in relationship to lifecycles and state changes
Interaction representation
Client management discussions
Deliver
Draft Detailed Management Arch for WSA Draft consideration
Proposal for base information
Management primer
Example Data: Service Environment
Identification information: Identifier, product name, version, installation date
Configuration information:
services deployed
log name
URL for runtime
Security settings for runtime
Metrics should be provided that indicate the rate of work and responsiveness of the execution environment to them
Start Time
Number of Requests
Number of Responses
Number of Failure Responses
Average Response Time of Responses
Average Response Time of Failure Responses
Average Response Time of Successful Responses
Operations: lifecycle management for itself and the services running in it:
start environment
stop environment
deploy a service
remove a service
Notifications: There are also a number of notifications that the execution environment can send:
service invocation failed
security access denied
State changes
Example Data: Web Service
Identification information:
service identifier
service name
service description
Configuration information:
access URL
WSDL description URL
configuration files
security settings
Metrics to help track usage of the service
Number of Service Requests, Responses, Failures
Average Service Response Time of Responses, Failures, Successful
Service Elapse Execution time
Number of Invocations, Failures Per Method
Average Response Time of Responses, Failures, Successes per Method
Total Elapsed Execution Time per Method
Operations to control service lifecycle
stop
start
Notifications:
state changes, service not found, service timed out, start failed, security failure, service failure, service not deployed, service unavailable, service failed, service deployed, service access denied
Example Data:Discovery Agency
Identification information:
product name,version,install date,maintenance level, instance name for instance, URL
Configuration information:
configuration file, replication period, nodes
Metrics: that will help operators gauge its responsiveness and usage (demarked with where the data must come from):
the rate of accesses, startTime
number of invokes per operation (get, find, publish, publish entity, publish relationship)
average response time per operation (get, find, publish etc)
percentage of request failure responses
summary metrics :# of businesses, services, relationships
Operations:
start, stop, smartPing, backup trace on/off, replicate (start, cancel, nodes, period), reset metrics, manageUser(find, manage, email)
Events: Degraded, running out of space, failure for percentage of messages,  overloaded runtime, about to fail, resource availability failure, has failed, resource failure, security access failures (decrypt, authentication, signature verification failures, authorization of users)