Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
change swapsyncd to module level. (#15501)
Description of PR Summary: change swapsyncd and disable_ipv6 to module level. Running time on T2 reduced by around 140 minutes (from 685.08 minutes to 543.95 minutes). Existing code will do swapSyncd for selected dut only in each iteration. In a T2 full test, it was around 8 LC which needs to do swapsyncd, including setup and teardown for each iteration. select_src_dst_dut_and_asic selected DUT to do swapsyncd (Existing) single_asic one downstream LC single_dut_multi_asic one downstream LC multi_dut_longlink_to_shortlink one upstream LC, one downstream LC multi_dut_shortlink_to_shortlink two downstream LC multi_dut_shortlink_to_longlink one upstream LC, one downstream LC After the fix, swapsyncd will be done for all LCs at the beginning of the test. Saved around 5 LC swapsyncd time select_src_dst_dut_and_asic selected DUT to do swapsyncd (after fix) Setup one upstream LC, two downstream LC single_asic none single_dut_multi_asic none multi_dut_longlink_to_shortlink none multi_dut_shortlink_to_shortlink none multi_dut_shortlink_to_longlink none Teardown one upstream LC, two downstream LC Type of change Bug fix Testbed and Framework(new/improvement) Test case(new/improvement) Back port request 202012 202205 202305 202311 202405 Approach What is the motivation for this PR? Reduce the run time for test_qos_sai module. How did you do it? change the swapsyncd fixture to module level. replace all dut with rpcsyncd container, instead of replacing selected dut multiple times for each iteration. How did you verify/test it? verified the physical testbed. co-authorized by: jianquanye@microsoft.com
- Loading branch information