Efficient way to write gitlab ci script

Question

I have a gitalb CI pipiline as follows.

Stage A

• job1
• job2
• job3

Stage B

• job1

In all the jobs in both stages, first package dependencies are installed. How can I handle this in an efficient way so that the total pipeline running time is the least. Also I can use multiple runners so job1, job2, job3 can run in parallel. Also multiple pipelines can run at a time.

Answer 1

You should define a cache: rule to cache your dependencies. Use an appropriate key: on the cache to ensure the cache is resorted at appropriate points.

See also: caching dependencies