Promote prefix_cache_retention_interval to an argument and change the default to 0 (#52216)

Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
This commit is contained in:
Tyler Michael Smith
2026-08-17 13:20:50 +00:00
committed by GitHub
co-authored by OpenAI Codex
parent 1d3a8b9e22
commit 017e9f4448
17 changed files with 160 additions and 63 deletions
+23
View File
@@ -476,6 +476,7 @@ def test_prefix_cache_default():
# should be None by default (depends on model).
engine_args = EngineArgs.from_cli_args(args=args)
assert engine_args.enable_prefix_caching is None
assert engine_args.prefix_cache_retention_interval == 0
# with flag to turn it on.
args = parser.parse_args(["--enable-prefix-caching"])
@@ -487,6 +488,28 @@ def test_prefix_cache_default():
engine_args = EngineArgs.from_cli_args(args=args)
assert not engine_args.enable_prefix_caching
args = parser.parse_args(["--prefix-cache-retention-interval", "64"])
engine_args = EngineArgs.from_cli_args(args=args)
assert engine_args.prefix_cache_retention_interval == 64
def test_prefix_cache_retention_interval_from_deprecated_env(
monkeypatch, caplog, disable_log_dedup
):
monkeypatch.setenv("VLLM_PREFIX_CACHE_RETENTION_INTERVAL", "64")
engine_args = EngineArgs()
assert engine_args.prefix_cache_retention_interval == 64
assert "VLLM_PREFIX_CACHE_RETENTION_INTERVAL" in caplog.text
assert "deprecated" in caplog.text
assert "prefix_cache_retention_interval" in caplog.text
parser = EngineArgs.add_cli_args(FlexibleArgumentParser())
args = parser.parse_args(["--prefix-cache-retention-interval", "32"])
engine_args = EngineArgs.from_cli_args(args)
assert engine_args.prefix_cache_retention_interval == 32
@pytest.mark.parametrize(
("arg", "expected", "option"),