panbingkun
panbingkun
> @panbingkun let's hold on this until we have a conclusion in #46600 @gengliangwang After #46600, the pr has updated.
In the `mima` step 1.with local maven repo cache: https://github.com/panbingkun/spark/actions/runs/9039162416/job/24841582106 https://github.com/panbingkun/spark/actions/runs/9039645041/job/24842838640 2.without local maven repo cache: https://github.com/panbingkun/spark/actions/runs/9040432971/job/24844609113 https://github.com/panbingkun/spark/actions/runs/9056537410/job/24879236204 https://github.com/panbingkun/spark/actions/runs/9057847717/job/24882503385
cc @LuciferYang @dongjoon-hyun
I have narrowed the scope of the impact of not using `local maven repo cache` to the step `MIMA test` of the job `lint`
> MIMA took 8 minutes and 6 seconds, and there doesn't seem to be a significant performance difference. I think this is feasible. > > Could you have it run...
I ran it twice again today. It was verified that the performance of the `mima` step seemed to be `stable` at about `8 min and ... seconds` without `local maven...
> This is fine for me, and the time is about the same as using the maven local cache. But I suggest splitting `disable maven local cache` and `upgrade to...
A test without `local maven repo cache` https://github.com/apache/spark/pull/46606
cc @gengliangwang
> LGTM overall. Thanks for the work! Updated, thanks! ❤️