More changes as part of the dualstack design:
- Change resolver and LB policy APIs to support multiple addresses per
endpoint. Specifically, replace `ServerAddress` with
`EndpointAddresses`, which encodes more than one address. Per-address
channel args are retained at the same level, so they are now
per-endpoint. For now, `EndpointAddress` provides a single-address ctor
and a single-address accessor for backward compatibility, so
`ServerAdress` is an alias for `EndpointAddresses`; eventually, this
alias and the single-address methods will be removed.
- Add an `EndpointAddressSet` class, which represents an unordered set
of addresses to be used as a map key. This will be used in a number of
LB policies that need to store per-endpoint state.
- Change the LB policy API's `ChannelControlHelper::CreateSubchannel()`
method to take the address and per-endpoint channel args as separate
parameters, so that we don't need to construct a legacy `ServerAddress`
object as we create a new subchannel for each address in the endpoint.
- Change pick_first to flatten the address list.
- Change ring_hash to use `EndpointAddressSet` as the key for its
endpoint map, and to use the first address of the endpoint as the hash
key.
- Change WRR to use `EndpointAddressSet` as the key for its endpoint
weight map.
Note that support for multiple addresses per endpoint is guarded in RR
by the existing `round_robin_delegate_to_pick_fist` experiment and in
WRR by the existing `wrr_delegate_to_pick_first` experiment.
This PR does *not* include support for multiple addresses per endpoint
for the outlier_detection or xds_override_host LB policies; those will
come in subsequent PRs.
Summary -
On the server-side, we are changing the point at which we decide whether
a method is registered or not from the surface to the transport at the
point where we are done receiving initial metadata and before we invoke
the recv_initial_metadata_ready closures from the filters. The main
motivation for this is to allow filters to check whether the incoming
method is a registered or not. The exact use-case is for observability
where we only want to record the method if it is registered. We store
the information about the registered method in the initial metadata.
On the client-side, we also set information about whether the method is
registered or not in the outgoing initial metadata.
Since we are effectively changing the lookup point of the registered
method, there are slight concerns of this being a potentially breaking
change, so we are guarding this with an experiment to be safe.
Changes -
* Transport API changes -
* Along with `accept_stream_fn`, a new callback
`registered_method_matcher_cb` will be sent down as a transport op on
the server side. When initial metadata is received on the server side,
this callback is invoked. This happens before invoking the
`recv_initial_metadata_ready` closure.
* Metadata changes -
* We add a new non-serializable metadata trait `GrpcRegisteredMethod()`.
On the client-side, the value is a uintptr_t with a value of 1 if the
call has a registered/known method, or 0, if it's not known. On the
server side, the value is a (ChannelRegisteredMethod*). This metadata
information can be used throughout the stack to check whether a call is
registered or not.
* Server Changes -
* When a new transport connection is accepted, the server sets
`registered_method_matcher_cb` along with `accept_stream_fn`. This
function checks whether the method is registered or not and sets the
RegisteredMethod matcher in the metadata for use later.
* Client Changes -
* Set the metadata on call creation on whether the method is registered
or not.
Revert the reversion of the SSL_CTX_new change (#34355 reverted #34180 )
with a fix.
There was an issue with using `strcpy` on a `new[] string` in the
constructor of `ssl_credentials`. An ASAN test caught this in some CI
down the line - `ERROR: AddressSanitizer: alloc-dealloc-mismatch
(operator new [] vs free)`
That `strcpy` call was changed to `grp_strdup` which duplicates a string
in a way that can be freed by `gpr_free` and should resolve the ASAN
failure.
This test assumed synchronous work serializer execution (or at least
faster async than we always get)... make a trivial change to keep the
test semantics but allow for the implementation to be more async.
Move the SSL_CTX to the level of the credentials rather than the
subchannel.
The SSL_CTX should only get created once per credential rather than once
per subchannel.
We should observe no behavior change with this PR, only efficiency
gains.
Most recent attempt was #34320, reverted in #34335.
The first commit here is a pure revert. The second commit fixes the
outlier_detection unit test to pass both with and without the
experiment.
This should address one of the failures we're seeing in #34224.
The test failure is caused by the changes in timing triggering a race
condition. In the code at head, we delay sending out the subscription
for the first CDS watch until we've already seen the other two CDS
watches, because the previous send_message op has not yet completed, and
by the time it does, we've seen all 3 watches, so we can send a
subscription for all 3 at the same time. With the WorkSerializer change,
the send_message op is complete by the time we see the first CDS watch,
so we subscribe to only that resource, and then later add the other two.
The result is that we'll NACK twice with two different messages, the
first one including only the error about the first resource, and the
second one including all three.
I suspect this same race condition would have been triggered eventually
by the EventEngine migration anyway; the current test basically depends
on the single-thread timing of the iomgr approach. So I'm addressing it
by replacing the e2e test with a unit test that covers the same cases
without the timing issue.
Rolls forward part of the dualstack changes, mostly from #33427 and a
little bit from #32692, both of which were reverted in #33718.
Specifically:
- For petiole policies, unconditionally start health watch on
subchannels, even if client side health checking is not enabled; in this
case, the health watch will report the subchannel's raw connectivity
state.
- Fix edge cases in health check reporting that occur when a watcher is
started before the initial state is reported.
- When client-side health checking fails, add the subchannel's address
to the RPC failure status message.
- Outlier detection now works only via the health checking watch, not
via the raw connectivity state watch.
- Remove now-unnecessary hack to ensure that outlier detection does not
work for pick_first.
Old message:
new_cluster_1: UNAVAILABLE: errors validating xds_cluster_resolver LB
policy config: [field:discoveryMechanisms error:must be non-empty]
New message:
new_cluster_1: FAILED_PRECONDITION: aggregate cluster graph has no leaf
clusters
Update from gtcooke94:
This PR adds support to build gRPC and it's tests with OpenSSL3. There were some
hiccups with tests as the tests with openssl haven't been built or exercised in a
few months, so they needed some work to fix.
Right now I expect all test files to pass except the following:
- h2_ssl_cert_test
- ssl_transport_security_utils_test
I confirmed locally that these tests fail with OpenSSL 1.1.1 as well,
thus we are at least not introducing regressions. Thus, I've added compiler directives around these tests so they only build when using BoringSSL.
---------
Co-authored-by: Gregory Cooke <gregorycooke@google.com>
Co-authored-by: Esun Kim <veblush@google.com>
CNR the flake, but I've changed the test (which is very old) to use some
of our more modern helper functions that have saner timeouts.
Also re-add a `return` statement that was accidentally removed in
#33753, which I noticed while working on this. Its absence doesn't cause
a real problem, but it does cause us to needlessly trigger a duplicate
connection attempt or report a duplicate CONNECTING update in some
cases.
De-experiment pick first since we have both affinity and randomness E2E
test running successfully.
---------
Co-authored-by: Yash Tibrewal <yashkt@google.com>
Fix sticky-TF behavior such that once we enter TRANSIENT_FAILURE, we do
not leave that state if we get a new address list.
Also, fix handling of subchannels in state TRANSIENT_FAILURE.
Previously, if a subchannel was already in state TRANSIENT_FAILURE when
we wanted to start a connection attempt on it (e.g., because the
subchannel already existed from a different channel, or because it
already existed in the previous subchannel list), we would wait for it
to report IDLE before attempting to connect. This PR changes pick_first
to instead immediately skip the subchannel and move on to the next one.
Now, the only time we wait for a subchannel in TRANSIENT_FAILURE is when
we wrap back around to the first subchannel in the list.
<!--
If you know who should review your pull request, please assign it to
that
person, otherwise the pull request would get assigned randomly.
If your pull request is for a specific language, please add the
appropriate
lang label.
-->
We decided to not populate `policy_name` with the HTTP filter name in
xDS case. So removing it from `GenerateServiceConfig`. This will be
consistent across languages. The gRFC
[PR](https://github.com/grpc/proposal/pull/346) has been updated.
I'm fairly certain that this path should be non-blocking (and making it
so makes the promise based code far more tractable).
This moves the blocking behavior into the blocking server_cc.cc function
that calls `grpc_server_shutdown_and_notify` instead of in that
non-blocking function.
---------
Co-authored-by: ctiller <ctiller@users.noreply.github.com>
This reverts the following PRs: #32692#33087#33093#33427#33568
These changes seem to have introduced some flaky crashes. Reverting
while I investigate.
As per gRFC A58, when WRR sees a subchannel report READY, it reset the
non_empty_since value, thus restarting the blackout period. However,
there were two cases where we were incorrectly triggering this code:
1. When WRR got an updated address list that contained addresses that
were already present on the old list and whose subchannels were already
in READY state, the initial notification for those subchannels on the
new list was READY, which incorrectly triggered resetting the
non_empty_since value.
2. Due to a bug in the outlier_detection policy, whenever an update was
propagated down through the OD policy without actually enabling OD, it
would incorrectly send a duplicate connectivity state notification for
the subchannels. This meant that a subchannel that was already in state
READY would report READY again, which would also incorrectly trigger
resetting the non_empty_since value.
This PR makes two changes:
1. Fix the bug in outlier_detection that caused it to generate the
spurious duplicate READY updates.
2. Fix WRR to reset the non_empty_since value when a subchannel goes
READY only if the subchannel has seen a previous state update and only
if that previous state was not READY. (The duplicate READY notifications
should not actually happen anymore now that the OD policy has been
fixed, but better to be defensive.)
Fixes b/290983884.
More work on the dualstack backend design:
- Now that all petiole policies have been changed to delegate to
pick_first, outlier detection no longer needs to eject via the
subchannel's raw connectivity state; it can now eject only via the
health state. See #33340.
- This also removes the now-unnecessary hack to explicitly disable
outlier detection in pick_first. See #33336.
More work on the dualstack backend design:
- Change ring_hash policy to delegate to pick_first instead of creating
subchannels directly.
- Note that, as mentioned in the WIP gRFC, because we lazily create the
pick_first child policies, so there's no need to swap over to a new list
as an atomic whole. As a result, we don't use the endpoint_list library
in this policy; instead, we just update a map in-place.
- Remove now-unused subchannel_list library.
More work on the dualstack backend design:
- Change round_robin to delegate to pick_first instead of creating
subchannels directly.
- Change pick_first such that when it is the child of a petiole policy,
it will unconditionally start a health watch.
- Change the client-side health checking code such that if client-side
health checking is not enabled, it will return the subchannel's raw
connectivity state.
- As part of this, we introduce a new endpoint_list library to be used
by petiole policies, which is intended to replace the existing
subchannel_list library. The only policy that will still directly
interact with subchannels is pick_first, so the relevant parts of the
subchannel_list functionality have been copied directly into that
policy. The subchannel_list library will be removed after all petiole
policies are updated to delegate to pick_first.
The address attribute interface was intended to provide a mechanism to
pass attributes separately from channel args, for values that do not
affect subchannel behavior and therefore do not need to be present in
the subchannel key, which does include channel args. However, the
mechanism as currently designed is fairly clunky and is probably not the
direction we will want to go in the long term.
Eventually, we will want some mechanism for registering channel args,
which would provide a cleaner way to indicate that a given channel arg
should not be used in the subchannel key, so that we don't need a
completely different mechanism. For now, this PR is just doing an
interim step, which is to establish a special channel arg key prefix to
indicate that an arg is not needed in the subchannel key.
The following bugs are fixed:
* Missing ExecCtx in event engine endpoints and listeners
* Ref counting issue with iomgr endpoint which causes crashes in
overloaded situations
The PR includes a test which triggers these bugs by simulating an
overloaded system.
Mark D. Roth
Yash Tibrewal
Eugene Ostroukhov
Gregory Cooke
Craig Tiller
nanahpang
jrandolf
Mohan Li
Yousuk Seung
Luwei Ge
Vignesh Babu