18 lines
1.1 KiB
Markdown
18 lines
1.1 KiB
Markdown
|
# Task Data
|
||
|
|
|
||
|
|
Task data is copied from task to task.
|
||
|
|
|
||
|
|
For example, consider the following sequence of tasks:
|
||
|
|
|
||
|
|
1. Start Event
|
||
|
|
2. Set Variable (sets variable a)
|
||
|
|
3. Read Variable (reads variable a)
|
||
|
|
|
||
|
|
In this scenario, if the "Set Variable" task sets the value of variable a to the number 1, the "Read Variable" task that follows can access that value (oh, cool, 1).
|
||
|
|
|
||
|
|
When the flow splits, as in a set of parallel branches, each branch gets a copy of the same data.
|
||
|
|
Then, when the data is merged back together at a merging gateway, if multiple branches write to the same variable, the last branch to complete will be the data that is propagated to tasks after the merging gateway.
|
||
|
|
Therefore, if you have tasks a, b, and c that run in parallel, it might not be a good idea for both a and b to set the "desired_withdrawal_amount" variable, since, depending on how fast a and b ran (and other implementation details around the parallel processing), the expected "winner" would not be knowable in advance.
|
||
|
|
|
||
|
|
This general branching and merging strategy is applied to all parallel constructs, including inclusive gateways.
|