transport&logistics services integration industry solution
TRANSCRIPT
At a time, when fuel prices have reached 30% of operating expenses for an average airline, there is no financial buffer room for an error in the schedule. At a time when personnel costs are a key factor in almost any industry, there is no room for crews to sit on stand-by or for people to wait while a vehicle or train goes through unscheduled maintenance. With customer satisfaction a key differentiator, proactive communication is a necessity, whether that means letting customers know where orders and packages are in the process or that their bag did not make their flight and how you will rectify the situation.
TRANSPORT & LOGISTICS INDUSTRY IT CHALLENGE
Myths endorsed by current process-driven Integration solutions
Reality: The Message-Driven ESB Effect
Adding a new service takes 2-4 weeks Zero-latency service modifications enabled by re-usability of message-driven Services
‘Certified Delivery’ of messages consumes valuable networking bandwidth
Peer-to-peer data transport obviates any penalties of networking bandwidth
Process changes necessitate service downtime and hence lost productivity
Dynamic service replacements without downtimes – no productivity disruptions
typically tied to particular platforms - Java or Microsoft’s .NET
Standards-based, interoperable with either Java or .NET platforms
Expensive solutions, often over $1,000,000
Equivalent benefits at a fractional cost, typically less than $200,000
MESSAGE-DRIVEN SOA
Flight data files need to be uploaded into their Flight Data Management applications, for every flight that lands and takes off at terminals.
These data files get saved to a file system from the flight computer as it gets hooked on to a loading gate, and also sent through emails and IBM MQSeries.
Each file needs to be transformed into a generic format, validated (which involves database lookup and other custom rules), enriched before its pushed into the target application.
SUBSET OF THE BUSINESS PROBLEM FOR ANALYSIS
Handling Crew Assignment Messages
contains details about the crew members allocated to a particular flight, such as name and contact information (e.g. address, phone number), qualifications, crew employer, start of duty information (e.g. time, airport), and flight specifications.
Each message needs to be transformed into a generic format, validated (which involves database lookup and other custom rules), enriched before its pushed into the one or more target application
SUBSET OF THE BUSINESS PROBLEM FOR ANALYSIS
Current solution suffers from the following issues, leading to delays in flight schedules.
• Data loss without a trace• limited visibility into the flow details. [for e.g parsing the positional
CPM data is a black box, its compiled code]• debugging is limited to log files• lack of error notification mechanism• Inherent performance issues - the flow is developed such it does
not scale simply by throwing in more hardware• For the business - IBM CrossWorlds/InterChange Server product
reached end of life -> meaning extra money on support.
PROBLEMS WITH CURRENT SOLUTION
STEP 1: RECEIVE CAM MESSAGES
CAM Messages come in 2 formats. Unsplit Flatfile format dropped into a specific directory in the file system and split XML CAM messages send on a queue in MQseries.
File Reader component picks up CAM messages from the filesystem and MQSeries component reads the CAM messages from MQseries queue.
File Reader
STEP 2: SPLIT CAM MESSAGE
CAMSplitter identifies flat file CAM messages in groups based on pre-determined criteria and split them into multiple such grouped
messages.
File Reader
STEP 3: GATE KEEPER PROCESSING
The XSLT component generates the UTC timestamp, determines the key information and creates a message envelope for the XML CAM messages.
This is done by Javascirpt component for the flat file CAM messages. Sequencing of the CAM messages is now performed by the Sequencer component for both the type of CAM messages.
File Reader
The flat file CAM messages are parsed using CAMParser, vaidated using CAMValidator and transformed using CAMTransformer components.
Similar parsing, validating and transformation is done using XSLT component for XML CAM messages.
STEP 4: PARSING AND CREATING CANONICAL XML
File Reader
A given CAM message may need to be routed to different destinations based on business rules determined dynamically.
The rules are injected to the DynamicCBR component that will apply the rule to the incoming CAM message. Based on the rule the DynamicCBR routes the CAM message to one or more destinations.
STEP 5: DESTINATION LOOKUP
File Reader
The CAM message that needs to be routed to the target destination is transformed to target format using the XSLT component
STEP 6: OUTPUT TRANSFORMATION
File Reader
Flatfile CAM messages are written to a target directory using FileWriter component and XML CAM is written to target queue on MQseies using MQseries component.
The sequencer needs to be notified about the completion of processing so that the next message in the corresponding group can be made available for processing.
STEP 7: DISPATCH TO TARGET SYSTEMS
File Reader
Reliability Transactional visibility Ease of use – lesser time to build the flows Cost of deployment
WHY FIORANO
Reliable data transfer – 100% guaranteed delivery Greater transactional visibility
• Runtime changes and debugging in one view Data and process high availability without the need for centralized
database Lesser time to develop the flows (Because of the Tools and the
Prebuilt components)• Ability to easily debug the flow of distributed events across multiple service
components• Ability to adapt to change dynamically
Separation of flow log and system log Equal citizen status to multiple programming languages
BENEFITS WITH FIORANO SOA PLATFORM