running this pipeline:

def do_i(p) {
return {
echo "$p"
for (i = 0; i < 10; i++)

{ echo "i: $i" }

}
}

stage('i'){
parallel(
"x" : do_i('xx'),
"y" : do_i('yy')
)
}

produces this output:
[Pipeline] stage
[Pipeline] {
[Pipeline] parallel
[Pipeline] [x] { (Branch: x)
[Pipeline] [y]

{ (Branch: y) [Pipeline] [x] echo [x] xx [Pipeline] [y] echo [y] yy [Pipeline] [x] echo [x] i: 0 [Pipeline] [y] echo [y] i: 1 [Pipeline] [x] echo [x] i: 2 [Pipeline] [y] echo [y] i: 3 [Pipeline] [x] echo [x] i: 4 [Pipeline] [y] echo [y] i: 5 [Pipeline] [x] echo [x] i: 6 [Pipeline] [y] echo [y] i: 7 [Pipeline] [x] echo [x] i: 8 [Pipeline] [y] echo [y] i: 9 [Pipeline] [y] }

[Pipeline] [x] echo
[x] i: 10
[Pipeline] [x] }
[Pipeline] // parallel
[Pipeline] }
[Pipeline] // stage
[Pipeline] End of Pipeline

Note how i is incremented as if both threads were referring to the same variable i.
If I modify the for loop to: for (int i = 0; i < 10; i++)
then it works as I expect, where each thread has its own i variable:

[Pipeline] stage
[Pipeline] {
[Pipeline] parallel
[Pipeline] [x] { (Branch: x)
[Pipeline] [y]

{ (Branch: y) [Pipeline] [x] echo [x] xx [Pipeline] [y] echo [y] yy [Pipeline] [x] echo [x] i: 0 [Pipeline] [y] echo [y] i: 0 [Pipeline] [x] echo [x] i: 1 [Pipeline] [y] echo [y] i: 1 [Pipeline] [x] echo [x] i: 2 [Pipeline] [y] echo [y] i: 2 [Pipeline] [x] echo [x] i: 3 [Pipeline] [y] echo [y] i: 3 [Pipeline] [x] echo [x] i: 4 [Pipeline] [y] echo [y] i: 4 [Pipeline] [x] echo [x] i: 5 [Pipeline] [y] echo [y] i: 5 [Pipeline] [x] echo [x] i: 6 [Pipeline] [y] echo [y] i: 6 [Pipeline] [x] echo [x] i: 7 [Pipeline] [y] echo [y] i: 7 [Pipeline] [x] echo [x] i: 8 [Pipeline] [y] echo [y] i: 8 [Pipeline] [x] echo [x] i: 9 [Pipeline] [x] }

[Pipeline] [y] echo
[y] i: 9
[Pipeline] [y] }
[Pipeline] // parallel
[Pipeline] }
[Pipeline] // stage
[Pipeline] End of Pipeline