proxy: Reduce round-trips to memcache and backend on info misses

Following a memcache restart in a SAIO, I've seen the following happen
during an object HEAD:

- etag_quoter wants to get account/container info to decide whether to
  quote-wrap or not
- account info is a cache miss, so we make a no-auth'ed HEAD to the next
  filter in the pipeline
- eventually this gets down to ratelimit, which *also* wants to get
  account info
- still a cache miss, so we make a *separate* HEAD that eventually talks
  to the backend and populates cache
- ratelimit realizes it can't ratelimit the request and lets the
  original HEAD through to the backend

There's a related bug about how something similar can happen when the
backend gets overloaded, but *everything is working* -- we just ought to
be talking straight to the proxy app.

Note that there's likely something similar going on with container info,
but the hardcoded 10% sampling rate makes it harder to see if you're
monitoring raw metric streams.

I thought I fixed this in the related change, but no :-/

Change-Id: I49447c62abf9375541f396f984c91e128b8a05d5
Related-Change: If9249a42b30e2a2e7c4b0b91f947f24bf891b86f
Related-Bug: #1883214
This commit is contained in:
Tim Burke 2023-02-28 23:03:23 -08:00
parent 2a90eadc0d
commit b68cc893f7
3 changed files with 40 additions and 2 deletions

View File

@ -438,6 +438,11 @@ def get_container_info(env, app, swift_source=None):
account = wsgi_to_str(wsgi_account)
container = wsgi_to_str(wsgi_container)
# Try to cut through all the layers to the proxy app
try:
app = app._pipeline_final_app
except AttributeError:
pass
# Check in environment cache and in memcache (in that order)
info = _get_info_from_caches(app, env, account, container)
@ -526,6 +531,11 @@ def get_account_info(env, app, swift_source=None):
account = wsgi_to_str(wsgi_account)
# Try to cut through all the layers to the proxy app
try:
app = app._pipeline_final_app
except AttributeError:
pass
# Check in environment cache and in memcache (in that order)
info = _get_info_from_caches(app, env, account)

View File

@ -152,7 +152,10 @@ class ZeroCacheDynamicResponseFactory(DynamicResponseFactory):
class FakeApp(object):
recheck_container_existence = 30
container_existence_skip_cache = 0
recheck_account_existence = 30
account_existence_skip_cache = 0
logger = None
def __init__(self, response_factory=None, statuses=None):
self.responses = response_factory or \
@ -352,6 +355,29 @@ class TestFuncs(BaseTest):
self.assertEqual([e['swift.source'] for e in app.captured_envs],
['MC', 'MC'])
def test_get_container_info_in_pipeline(self):
final_app = FakeApp()
def factory(app):
def wsgi_filter(env, start_response):
# lots of middlewares get info...
if env['PATH_INFO'].count('/') > 2:
get_container_info(env, app)
else:
get_account_info(env, app)
# ...then decide to no-op based on the result
return app(env, start_response)
wsgi_filter._pipeline_final_app = final_app
return wsgi_filter
# build up a pipeline
filtered_app = factory(factory(factory(final_app)))
req = Request.blank("/v1/a/c/o", environ={'swift.cache': FakeCache()})
req.get_response(filtered_app)
self.assertEqual([e['PATH_INFO'] for e in final_app.captured_envs],
['/v1/a', '/v1/a/c', '/v1/a/c/o'])
def test_get_object_info_swift_source(self):
app = FakeApp()
req = Request.blank("/v1/a/c/o",

View File

@ -513,7 +513,8 @@ class TestController(unittest.TestCase):
def test_get_account_info_returns_values_as_strings(self):
app = mock.MagicMock()
app._pipeline_final_app.account_existence_skip_cache = 0.0
app._pipeline_final_app = app
app.account_existence_skip_cache = 0.0
memcache = mock.MagicMock()
memcache.get = mock.MagicMock()
memcache.get.return_value = {
@ -539,7 +540,8 @@ class TestController(unittest.TestCase):
def test_get_container_info_returns_values_as_strings(self):
app = mock.MagicMock()
app._pipeline_final_app.container_existence_skip_cache = 0.0
app._pipeline_final_app = app
app.container_existence_skip_cache = 0.0
memcache = mock.MagicMock()
memcache.get = mock.MagicMock()
memcache.get.return_value = {