100 likes | 251 Views
Workflow has moved on …. Charles Brett. Objective and structure. Look at background to work flow messaging application broking process flow automation (PFA) work flow automation (WFA) Unresolved issues Other thoughts/dimensions
E N D
Workflow has moved on … Charles Brett
Objective and structure • Look at background to work flow • messaging • application broking • process flow automation (PFA) • work flow automation (WFA) • Unresolved issues • Other thoughts/dimensions • Introduce Kevin Hudson and Dale Skeen – to explore work flow and B2B developments in more detail
Messaging • First there was ‘internal messaging’ • Next TCAM and IMS (DC) • Then there was little (except S.W.I.F.T./EDIFact, etc.) • Then MQSeries, TIB, etc. • Then AQ • Now JMS
Application brokers (aka message brokers, integration brokers, etc.) • Broking between existing (and new) applications (minimally invasive) • Primarily Intra-business (Intra-B) • Usually based on messaging • added ‘intelligence on top of messaging (primarily queuing) • added cottage industry: ‘transformation’, ‘routing’, etc. • added ‘external engine’ as convergence mechanisms • Limited transaction and database awareness
Rise of … • Then the following arrived: • need for co-ordination above transformation/routing (process to process) • XML/RosettaNet (and PIPs), as agreed ‘media’ • Inter-business opportunities (Inter-B or B2B) • re-awakening of transaction interest, especially in context of long running activities within and between businesses • Re-awakening of database involvement
Process flow automation (PFA) and work flow automation (WFA) • PFA • a means to manage/co-ordinate processes (not people) • built as super-set on/over application brokers/messaging; implemented as external engine • events (and Event/Action/Condition processing) to the fore – allegedly business oriented • WFA • added people component, especially for actions/decisions/confirmation • Increasingly implemented as external engine, co-ordinating across multiple existing applications • Divergent from traditional ‘workflow’ – which was an application; WFA is an engine working to manage other applications
Unresolved issues (all variants on 2 Generals problem) • Security • Trust • Acknowledgement • Non-repudiation • Audit/Audit tracking • Ownership/Responsibility • Legality • Long running activities • Business activity analysis
Other thoughts/dimensions • Queues were flat files, not tables in databases • Messages carried data payload; now carry XML as well • Transaction awareness forgotten/omitted • Development/Operations awareness divergence • The mythical ‘Business Analyst’ • The implacable growth of external - to application - ‘ancillary’ systems (applications brokers, PFA, WFA and even applications servers) • ‘Business event’ handling re-appears yet again (last time was OO-led (retarded)
Net position • Traditional work flow has moved on, and bifurcated • Messaging and queuing taken as read (no longer interesting – although poor variety of choice) • Transaction and database importance is on the rise (yet again); but poorly understood • Intermediate ‘mechanism engines’ on the increase • PFA is where IT interest lies: ‘about systems not people’ • WFA not an application, more of a manager; but people dimension is a complication • Workflow still exists (primarily paper trail oriented) • Operational implications have had little thought
For the real insights into the technologies have advnced • Kevin Hudson (Senior Product Manager, Oracle) • advances in WFA • Dale Skeen (CTO, Vitria plus a co-founder of TIBCO) • work flow and the B2B dimension