chore: ignore boot test resume_time metric in A/B

This metrics has very small numbers which makes it too
volatile for A/B tests.

Signed-off-by: Egor Lazarchuk <yegorlz@amazon.co.uk>

Author: ShadowCurse

tests/framework/ab_test.py

@@ -160,7 +160,7 @@ def git_ab_test_host_command(
 
 
 def set_did_not_grow_comparator(
-    set_generator: Callable[[CommandReturn], set]
+    set_generator: Callable[[CommandReturn], set],
 ) -> Callable[[CommandReturn, CommandReturn], bool]:
     """Factory function for comparators to use with git_ab_test_command that converts the command output to sets
     (using the given callable) and then checks that the "B" set is a subset of the "A" set

tools/ab_test.py

@@ -49,6 +49,8 @@
     # block latencies if guest uses async request submission
     {"fio_engine": "libaio", "metric": "clat_read"},
     {"fio_engine": "libaio", "metric": "clat_write"},
+    # boot time metrics
+    {"performance_test": "test_boottime", "metric": "resume_time"},
 ]