The effective layering of ResponseWriters is today, from outside to
inside, httplog(timeout(audit(metrics(original)))). From
6e3fd91e1aa3259d7bd67e0a65693e346ade347d, calls to http.Error in the
apiserver's root healthz handler use an unwrapped ResponseWriter --
effectively timeout(audit(metrics(original))) -- to avoid logging
stack traces for those requests.
From 0d50c969c587c8a6c16e0962118305ac652c5a6b, the same call to
http.Error receives a completely-unwrapped ResponseWriter. This has
the effect of bypassing not only the httplog wrapper, but also
timeout, audit, and metrics. The timeout wrapper defends against
the (disallowed) use of underyling ResponseWriter after the completion
of its request's ServeHTTP call. Since that defensive behavior is
being bypassed, it's possible for the root healthz handler to panic
when health probes time out.
Instead of continuing to use a wrapper-aware means of disabling stack
traces, this commit adds a new function to httplog that allows
customization of the stack trace logging predicate on a per-request
basis.
Kubernetes-commit: ff849fe8b688606d5173d5ee0213a96cffae23c0
In 18177e2bdeafbddeb3d66fec0b8cb88794cd69ff, PostGoRestful field has
been renamed to NonGoRestfulMux, but the documentation change did not
follow.
This commit fixes that to avoid potential confusion.
Signed-off-by: Mateusz Gozdek <mgozdek@microsoft.com>
Kubernetes-commit: 36ede8de945adcc06bde21c49dc157e9c741e0d2
This PR removes Serve function and uses all required places
ServeWithListenerStopped which takes place new Serve function.
This function returns ListenerStopped channel can be used to drain
requests before shutting down the server.
Kubernetes-commit: a8d2b3a7926394b1c53621804cdeb93e4a61b7c8
Print information from both the original path registration and the new
path registration stack traces when encountering a duplicate. This helps
the developer determine where the duplication is coming from and makes
it much easier to resolve.
Signed-off-by: Andy Goldstein <andy.goldstein@redhat.com>
Kubernetes-commit: 04aa8f9dcdbc575fde37e25e45315359b0aa1ca6
Implements server side field validation behind the
`ServerSideFieldValidation` feature gate. With the
feature enabled, any create/update/patch request
with the `fieldValidation` query param set to
"Strict" will error if the object in the request
body have unknown fields. A value of "Warn"
(also the default when the feautre is enabled)
will succeed the request with a warning.
When the feature is disabled (or the query param
has a value of "Ignore"), the request will succeed
as it previously had with no indications of any
unknown or duplicate fields.
Kubernetes-commit: e50e2bbc889eb274ad1463a54188a2805767bfde
addresses feedback by adding a test to make sure that the manager pipeline will not return the same instance used as input, and that the output does not input managedFields
Kubernetes-commit: 68e175ef68eff1351917ac206fb29abe42128062
drop the managed fields of the objects from the audit entries when we
are logging request and response bodies.
Kubernetes-commit: bbc59348318c29199e23b27981fb56436ac68705
Split process() function into processEvents() and process().
This is done in anticipation of GetAllEventsSinceThreadUnsafe()
returning an entity using which events can be constructed and
not the events itself.
Subsequently, this commit also moves updating resource version
for initEvents from Watch() to the processEvents() func.
Signed-off-by: Madhav Jivrajani <madhav.jiv@gmail.com>
Kubernetes-commit: aab7cd3d8a66f425022ca5b2a2bd0d3019efe526
the new field exists primarily to avoid returning a 404 response when a resource actually exists but we haven't installed the path to a handler.
it is exposed for easier composition of the individual servers.
the primary users of this field are the WithMuxCompleteProtection filter and the NotFoundHandler.
Kubernetes-commit: ddfbb5d2bb57ee44b3e10f0b58f9cc7001f55802
It puts the muxCompleteProtectionKey in the context if a request has been made before muxCompleteSignal has been ready.
Putting the key protect us from returning a 404 response instead of a 503.
It is especially important for controllers like GC and NS since they act on 404s.
The presence of the key is checked in the NotFoundHandler (staging/src/k8s.io/apiserver/pkg/util/notfoundhandler/not_found_handler.go)
The race may happen when a request reaches the NotFoundHandler because not all paths have been registered in the mux
but when the registered checks are examined in the handler they indicate that the paths have been actually installed.
In that case, the presence of the key will make the handler return 503 instead of 404.
Kubernetes-commit: b71fa61b79598b723c3ee23217e0b44564d90b52
The new handler is meant to be executed at the end of the delegation chain.
It simply checks if the request have been made before the server has installed all known HTTP paths.
In that case it returns a 503 response otherwise it returns a 404.
We don't want to add additional checks to the readyz path as it might prevent fixing bricked clusters.
This specific handler is meant to "protect" requests that arrive before the paths and handlers are fully initialized.
Kubernetes-commit: 53867975e72c7a2d2dd94aac6bd2869411f92094
Track the introduction of FinalSeats.
Give up on calculating expected results for tests with added latency,
because I did not find an easy and obvious way to do it.
Kubernetes-commit: 0fc595e03360ba7fc4c3e251d4b41f39172aca72
New anti-windup technique: use the request arrival time as the floor
on the virtual dispatch time. Prevent bound violations where they
might arise rather than fixing up just one queue at dispatch time,
so that the fixed up dispatch times figure into the dispatching choice.
Two tweaks to the shuffle sharding. Take seats of executing requests
into account as well as seats of waiting requests. Do not always
consider the generated hand in the same order.
Rename the queueset methods that do shuffle sharding and finding the
queue to dispatch from, because the old names were confusingly
similar.
Tighten up some request margins.
Name the test cases in TestNoRestraint and TestWindup.
Kubernetes-commit: 4b9cba85874158b25b5c994773a4ec04343820c2
Canonicalize listing of test cases.
Make TestNoRestraint try both cases: competition and none.
Kubernetes-commit: 0ee1a7b4ff9012b050bd447055ad5e1e8c57c30e
Make TestNoRestraint verify that fairness is NOT achieved
when there is real competition.
Make TestWindup run two cases, to show that 0.1 is too narrow
a margin and 0.26 is wide enough.
Kubernetes-commit: c4945fdf0c14ba2032a5c8edf192678d9fe00374
This PR enables unaryClientInterceptor in conjunction with Prometheus interceptor.
Previously it was simply overwritten by the Prometheus interceptor.
As a result etcd client didn't attempt to retry certain errors.
The unaryClientInterceptor is important because it knows how to retry all sorts of errors from the etcd cluster. It will make the API server more resilient to failures - end users won't see certain errors.
The full list of retriable (codes.Unavailable) errors can be found at https://github.com/etcd-io/etcd/blob/main/api/v3rpc/rpctypes/error.go#L72
Kubernetes-commit: 83171562b0954b2e19eb69943f01a44779cc7a8f
These behavioral unit tests of queueset were failing because the
evaluation criteria were too strict.
Kubernetes-commit: 59d319ec06bb33289a87036418b4a61ed3bb215f
So that the width estimate has some effect but not a grossly excessive
one.
Added the fifo::Peek method to simplify the fifo client code.
Also renamed the queueSet::estimatedServiceTime field to
estimatedServiceSeconds to make the units clear.
Kubernetes-commit: a0c161f2f6908ee424ea888ff40f75ff071bd20a
Instead of a plain `Mutex`, use an `RWMutex` so that the common
operations can proceed in parallel.
Kubernetes-commit: 58927c1abede11ce7a8a74104328cf823df1b39e
The cmp comparison is relatively expensive (#104821). If we're not
going to log it, we shouldn't make the comparison.
Kubernetes-commit: f9f556dc7061df1dfc8c1628db983eeb97149317
This is a Config specialized for a GroupResource.
It will support generating new resource-specific metrics.
Kubernetes-commit: 85bcd243aa3c8769a5904a1aea44ce704f5e7174
This reverts commit fc5863b8b276e0789f717859e8cce58d7d060181, reversing
changes made to 027fe2554fd18343b8be39eddc8ff6570a6c390f.
Kubernetes-commit: f9f08725907b7db2104ee5fe9f82ab0752726533
This reverts commit edb0a72cff0e43bab72a02cada8486d562ee1cd5, reversing
changes made to 80feff6f407be9f0898c449ba3f9d4d013f05ec9.
Kubernetes-commit: 8844d3092a46a40915b4df6e3b9944d5081f8268
Now tries a little harder to get to meaningful stack frames.
This revision seems to strike a balance that is useful in the queueset tests.
Kubernetes-commit: b56dd725032cb7a14aa27e4c50c1c9d7c6d23eb1
If one doesn't DeepCopy() on the way into Create, we can end up writing
into the original object. This is by design, and should not be a
problem EXCEPT for tests. If a test compares the input to this function
with the result, but the input was mutated in-situ, it may hide errors,
resulting in tests that pass, but shouldn't.
Kubernetes-commit: 6dfae64d9bebb2c40680bbd6e8270f69839ab013
Jordan Liggitt