tests: make size comparisons more lenient (#9388)

The empirically determined threshold doesn't hold for PG 17.
Bump the limit to stabilise ci.
This commit is contained in:
Vlad Lazar
2024-10-14 16:50:12 +01:00
committed by GitHub
parent d92ff578c4
commit f4f7ea247c

View File

@@ -479,9 +479,9 @@ def assert_size_approx_equal(size_a, size_b):
"""
# Determined empirically from examples of equality failures: they differ
# by page multiples of 8272, and usually by 1-3 pages. Tolerate 4 to avoid
# by page multiples of 8272, and usually by 1-3 pages. Tolerate 6 to avoid
# failing on outliers from that observed range.
threshold = 4 * 8272
threshold = 6 * 8272
assert size_a == pytest.approx(size_b, abs=threshold)