Shulou(Shulou.com)06/02 Report--

DatwWorks finally upgraded 2.0, very excited, but also a little worried. Because there is no way to create a new and old workflow.

The new version abandons the concept of "workflow" and introduces two new concepts, "business process" and "solution". So, as a team Leader, I spent a lot of energy delving into how these two concepts should be applied to my actual business.

1. Purpose

Can apply many powerful new functions of DataWorks2.0 as soon as possible

Solve the inconvenience caused by the inability to create a new old workflow as soon as possible

two。 Old workflow division

Task nodes are generally divided into the following categories:

Source data synchronization node single business basic data layer statistics node cross-business basic data layer statistics node result data layer statistics node result data synchronization node 2.1 partition dimension

The old version of workflow mainly consists of the following dimensions:

2.1.1 in which database does the source database data source include: source data synchronization node 2.1.2 product line source data belongs to which product line includes: source data synchronization node + single business basic data layer statistics node 2.1.3 which system is the target system result data provided to (background system, BI, DataV, …... ) including: result data layer statistics node + result data synchronization node 2.1.4 which business module does the result data belong to (order + complaint, order + insurance, merchant + order, …... ) includes: cross-business basic data layer statistics node + result data layer statistics node + result data synchronization node 2.1.5 operation cycle daily, weekly and monthly operation includes: all nodes 2.1.6 input / statistics / output contains the types of nodes Whether it belongs to "source data migration", "statistical task" or "result data migration" includes: all nodes 2.2 other partition principles minimize the number of upstream workflows on which each workflow depends Can give the whole workflow a name in one sentence; the number of nodes in each workflow is no more than 30; 3. The difference between the old workflow and the new business process 3.1 run cycle the old workflow:

The entire workflow (including multiple internal nodes) corresponds to a running cycle of the new business process:

Each node corresponds to a run-time 3.2 dependency

Legacy Workflow:

Workflow depends on workflow; internal nodes depend on internal nodes

New version of Business process:

There is no dependency between business processes; nodes depend on nodes, which can be relied on across business processes and between tasks with different running cycles. 3.3 Node Classification Legacy Workflow:

There is no classification, and the new version of business process can only be distinguished by the naming of internal nodes:

Can be automatically classified by node type (SQL node, data synchronization node, machine learning node,... ) 3.4 Classification of workflows (business processes) legacy workflows:

There is no classification, and the new version of business processes can only be distinguished by the naming of workflows:

It can be classified by the "solution" function (similar to tagging each business process), and the "business process" and "solution" are many-to-many relationships. 3.5 submit scope legacy workflow:

The entire workflow (including multiple internal nodes) submits the new version of the business process together:

A single task node is submitted. It also supports the batch submission of multiple nodes in the business process. 3.6 rename

Legacy Workflow:

Workflow does not support renaming; internal nodes do not support renaming

New version of Business process:

The business process supports renaming; the node supports renaming; 3.7 add the location of the folder to the old workflow:

Multi-level folder → Workflow → Internal Node New version of Business process:

Business process → data integration / data development → multi-level folder → node 4. Principles of business process division

Finally, with regard to the division principles of my business process, the summary is as follows:

Business processes complete a minimum business; the number of business processes can be slightly more, which can be found through the solution; it is best not to have too many nodes in each business process, and only so many can be visually displayed in each business process.

I hope you can use many of the powerful functions of DataWorks2.0 as soon as possible! You don't have to sew and mend in the old workflow.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.