Is it platform specific
generic
Importance or Severity
Critical
Description of the test gap
GCU has no test that verifies apply-patch completes within a reasonable
time when processing config changes. GCU performance regressions — including
near-hung behavior — go completely undetected in CI.
Known repro:
- ACL table configured with ports bound to a large list of interfaces
- GCU patch replaces the ports field with a very short list
- GCU internal sort/diff logic hangs or takes 60+ minutes
Affected Area
GCU (Generic Config Updater) — tests/generic_config_updater/
Impact of missing tests and proposed tests
A build where GCU hangs passes all current tests because no test asserts
timing. The regression only surfaces in production.
Proposed fix: new test test_gcu_acl_ports_replace_large_to_small in
tests/generic_config_updater/test_gcu_acl_ports_replace.py that:
- Creates ACL table bound to all available ports
- Replaces ports field with 2 ports via GCU patch
- Asserts patch completes within get_gcu_timeout(duthost)
This test FAILS on a buggy build and PASSES after the fix.
Is it platform specific
generic
Importance or Severity
Critical
Description of the test gap
GCU has no test that verifies apply-patch completes within a reasonable
time when processing config changes. GCU performance regressions — including
near-hung behavior — go completely undetected in CI.
Known repro:
Affected Area
GCU (Generic Config Updater) — tests/generic_config_updater/
Impact of missing tests and proposed tests
A build where GCU hangs passes all current tests because no test asserts
timing. The regression only surfaces in production.
Proposed fix: new test test_gcu_acl_ports_replace_large_to_small in
tests/generic_config_updater/test_gcu_acl_ports_replace.py that:
This test FAILS on a buggy build and PASSES after the fix.