Add an http filter to extract the original source on a per-http-request basis
Signed-off-by: Kyle Larose <kyle@agilicus.com>
Mirrored from https://github.com/envoyproxy/envoy @ 712a72a718969ef37eb05a57a7e11c94d5fe6253
Implements the hedge_on_per_try_timeout option in HedgePolicy config
which instructs the router filter to not cancel requests that hit the
per try timeout before sending a retry. This means the router must be
able to manage multiple simultaneous upstream requests and handle
deciding which response "wins" and canceling any other in flight
requests.
Fixes#5841
Signed-off-by: Michael Puncel <mpuncel@squareup.com>
Mirrored from https://github.com/envoyproxy/envoy @ 4d4748ea243e0ab0538bcb0894a64bb0da617b55
Implement the scoped RDS (SRDS) API config subscription and provider based on the config protos introduced in #6675 and the ConfigProvider framework introduced in #5243 and #6781.
NOTES:
See parent PR #5839 for full context into these changes. PRs 2a (#6675) and 2b (#6781) have already been merged.
The API is not yet fully implemented. This PR introduces static and dynamic (xDS config subscription) handling of scoped routing configuration, but the new L7 multi tenant routing logic (see #4704) has not yet been introduced.
The API is not yet plumbed into the HttpConnectionManager, that will be done in the next PR.
This PR includes unit tests only; integration tests will follow in the next PR.
Risk Level: Low (this DS API is not yet integrated into the HCM and can not be enabled via config).
Testing: Unit tests added.
Docs Changes: N/A.
Signed-off-by: Andres Guedez <aguedez@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 831d0cbb389f241fd4d1a8b224ffa43e4257a680
Removes golang/protobuf dependency. This is an old @sesmith177 TODO and now the version of rules_go we import has the requisite well known types (WKT). Bumping rules_go to the latest version 0.18.4 (release notes) aligns with PGV which is on 0.18.3.
Risk Level: low (I think but there are WKT changes)
Testing: bazel test //test/...
Signed-off-by: Michael Payne <michael@sooper.org>
Mirrored from https://github.com/envoyproxy/envoy @ 7bf53149d137dcb2a7b4eddb4f8c23eda23d66e5
The TLS session ID is now accessible as a method on the Ssl::ConnectionInfo object.
Add a new %DOWNSTREAM_TLS_SESSION_ID% directive to the file access logger, and a
tls_session_id field to the gRPC access logger. Update unit tests to validate the
new access log fields.
Signed-off-by: Mike Grass <mgrass@salesforce.com>
Mirrored from https://github.com/envoyproxy/envoy @ cbd8d57e9fc2921d26236db24b6bf91246ba3b4f
Risk Level: none, adding unused proto field
#4991
Signed-off-by: Fred Douglas <fredlas@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 154cdcb25832b01794430bc473e6675d365dd082
Changes: Fixes coverage test hang.
No change to core logic from #6692.
Signed-off-by: Karthik Reddy <rekarthik@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 27b3e48e8d754dcb956dca58888c7dbfde7ceec8
This adds an option to allow hosts to be excluded in lb calculations until they have been health checked
for the first time. This will make it possible to scale up the number of hosts quickly (ie large increase
relative to current host set size) without triggering panic mode/spillover (as long as the initial health check
is succeeds).
While these hosts are excluded from the lb calculations, they are still eligible for routing when panic
mode is triggered.
Signed-off-by: Snow Pettersen <snowp@squareup.com>
Mirrored from https://github.com/envoyproxy/envoy @ 4c80194bf82193261aa52a4ca64c4e6a461881c0
Moved all of DeltaSubscriptionImpl's existing tests to be DeltaSubscriptionState tests, since that's what they really were. Added tests that test actual DeltaSubscriptionImpl logic. Followup of #6729.
Risk Level: none, just adding tests
Testing: DeltaSubscriptionState, DeltaSubscriptionImpl
Signed-off-by: Fred Douglas <fredlas@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ ae3106b3ee792f697b2098781a0e65d8d73a32ca
Update buildifier from 0.22.0 to 0.25.0 (release notes), bazel-compilation-database from 0.3.2 to 0.3.3 (release notes), bazel-skylib from 0.7.0 to 0.8.0 (release notes) and change to developer provided tar download and protoc-gen-validate from 0.13.0 to 0.14.0 (release notes).
Risk Level: Low
Testing: bazel bazel test //test/...
Docs Changes: none required
Release Notes: none required
Signed-off-by: Michael Payne <michael@sooper.org>
Mirrored from https://github.com/envoyproxy/envoy @ 8b1f6eaa7aa28c14640b2ed779738a8d7be33a34
Description: PR #6211 updated the documentation of CheckResponse.status to reflect Envoy's actual behavior at the time. Later, PR #6505 changed that behavior to be in-line with the pre-6211 docs. So, revert that part of PR #6211.
Risk Level: Low
Testing: None
Docs Changes: Inline in API protos
Release Notes: none
Signed-off-by: Luke Shumaker <lukeshu@datawire.io>
Mirrored from https://github.com/envoyproxy/envoy @ 5fef1519f4ce3fa2dd1389bfa390e7294182b1e7
As a first step towards TDS, add support for proto/JSON representations of the runtime virtual file
system and plumb this into bootstrap.
"numerator" and "denominator" are now reserved keywords and can't be used in runtime keys; this
simplifies parsing significantly. It is not expected that these are widely used, due to the
difficulty of ensuring atomicity if representing fractions, but this is technically a runtime API
breaking change.
Risk level: Medium (due to the reserved word changes).
Testing: Unit and server tests added.
Part of #6708
Signed-off-by: Harvey Tuch <htuch@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ eb699ab2d53aadf2c5a26ed02fcc1e7de1df7d8a
Signed-off-by: Adam Kotwasinski <adam.kotwasinski@gmail.com>
Mirrored from https://github.com/envoyproxy/envoy @ ecf1085a5b279d1244a02433f204ebba5306d40f
Fix a build warning:
INFO: From ProtoGenValidateCcGenerate external/envoy_api/envoy/service/auth/v2/attribute_context.pb.h:
envoy/service/auth/v2/attribute_context.proto: warning: Import envoy/api/v2/core/base.proto but not used.
Signed-off-by: Ismo Puustinen <ismo.puustinen@intel.com>
Mirrored from https://github.com/envoyproxy/envoy @ 16e85b1ffc4fceb3c1c3857b96862937126468a8
Description: Adding API support to report Locality proximity information as part of Locality Assignment. This does not need to be documented yet, and should be documented as part of LB algorithm implementation which will use this information.
Risk Level: LOW
Testing: None
Signed-off-by: Vishal Powar <vishalpowar@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 5e039a6581cf5fa1f7b70329afeb2e3d35b8189c
When using active health checking, hosts are not removed from
dynamic clusters if they are still passing health checks. This
creates a situation in which hosts might not be removed for a
very long time if the sequence is reversed; removal followed by
health check failure. This change handles the second case so that
any time a host is both removed AND failing active health check,
in any order, it will be removed.
This has been an issue "forever" but is more obvious when using
streaming EDS or very long polling DNS.
Fixes https://github.com/envoyproxy/envoy/issues/6625
Signed-off-by: Matt Klein <mklein@lyft.com>
Mirrored from https://github.com/envoyproxy/envoy @ 41eefffcd728d071037a57a1accd402ec188bcd5
total_issued_requests field tracks the count of requests issued since
the last report. This field will be used for global load balancing
decisions.
Envoy maintains a stats counter rq_total to track total
requests made. By latching this counter when load reporting period
begins, we are able to count the total requests issued in a load
reporting interval. This information is then used to populate the field
total_issued_requests.
Risk Level: Low
Testing: //test/integration:load_stats_reporter passes.
Signed-off-by: Karthik Reddy <rekarthik@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ cdddf54e8548036703ee928335a44c3b503b988b
Protos to statically and dynamically (via the Scoped Route Discovery Service) configure scoped
routing logic in Envoy.
Signed-off-by: Andres Guedez <aguedez@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 2778c3c0c1e1f18a1c7864b4e7707e9d0ca2571e
It's helpful to a client-side LB to know the total endpoint serving QPS when making LB decisions.
This could also be used as a signal for Envoy's least loaded LB.
Signed-off-by: Harvey Tuch <htuch@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 0d5bfe2db00075d12b163b27f130bbca8539ffbc
Dynamic link tests in ASAN to save disk space. Partially addresses #1407.
Risk Level: Low (ASAN only)
Testing: CI
Signed-off-by: Lizan Zhou <lizan@tetrate.io>
Mirrored from https://github.com/envoyproxy/envoy @ 8cca5c2992693ae143834ad0726aa630058f5910
Signed-off-by: Brian Rectanus <brectanus@signalsciences.com>
Mirrored from https://github.com/envoyproxy/envoy @ 77d267eef63edc215f93d140f61bfc711b59a31b
Reintroduce dispatcher stats previously reverted in #6649. Dispatcher stats are now gated on a new bootstrap config parameter, enable_dispatcher_stats, and disabled by default.
Risk Level: Low
Testing: Manually verified that config is respected (see #6582, this still requires integration tests).
Docs Changes: Included a note about how statsd doesn't play well with these high-volume histograms.
Signed-off-by: Dan Rosen <mergeconflict@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 481a127f5ade32e5590b4e8b8271576e44643082
This adds support for modifying the grpc-timeout provided by the
downstream by some offset. This is useful to make sure that Envoy is
able to see timeouts before the gRPC client does, as the client will
cancel the request when the deadline has been exceeded which hides the
timeout from the outlier detector.
Signed-off-by: Snow Pettersen <snowp@squareup.com>
Mirrored from https://github.com/envoyproxy/envoy @ 03ae1ef6b9afcfda9545b3d734b54027776254f9
realized that, with the unreliable queue implementation copied from SotW xDS, delta xDS could get into a state where Envoy thinks it has subscribed, but the server hasn't heard the subscription, with no way for either to realize the mistake. I fixed that by converting the queue setup to a cleaner "do I currently want to send a request?" with the request's (un)subscriptions only populated immediately before the request is actually sent into gRPC.
While doing that, I further realized there was a problem when a given resource was subscribed then unsubscribed (or reversed), all in between request sends. I made sure Envoy handles that sensibly, and added explicit requirements to the xDS protocol spec to ensure servers will also handle it sensibly.
Added unit tests for those fixes.
Risk Level: low
Testing: added unit tests for bugs uncovered
#4991
Signed-off-by: Fred Douglas <fredlas@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 3abb7d051c55f190d19b531a2fd1360842b334a5
Default behavior remains unchanged: retries will use the runtime parameter
defaulted to 25ms as the base interval and 250ms as the maximum. Allows
routes to customize the base and maximum intervals.
Risk Level: low (no change to default behavior)
Testing: unit tests
Doc Changes: included, plus updated description of back-off algorithm
Release Notes: added
Signed-off-by: Stephan Zuercher <zuercher@gmail.com>
Mirrored from https://github.com/envoyproxy/envoy @ 5ea103ee0e88cca52f543fdd230ca9f84c518c88
Created OpenRCA service proto file based on ORCA design
Risk Level: Low
Signed-off-by: Chengyuan Zhang <chengyuanzhang@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ 5cb222964fb691d8cb6cb0f6dd8fdd7b5dbdbb61
Add support for specifying _stale_after timeout as part of ClusterLoadAssignment
Risk Level: Low
Optional Feature that is triggered by the Management Server. Defaults to noop.
Testing: Unit test
Docs Changes: None
Release Notes: None
Fixes#6420
Signed-off-by: Vishal Powar <vishalpowar@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ e6a3ce16ff818cc4d5588345fca084efeaa1985d
Change the behavior of the delayed_close_timeout such that it won't trigger unless there
has been at least a delayed_close_timeout period of inactivity after the last write event on
the socket pending to be closed.
This mitigates a race where a slow client and/or low timeout value would cause the socket
to be closed while data was actively being written to the socket. Note that this change does
not eliminate this race since a slow client could still be considered idle by the updated timeout
logic, but this should be very rare when useful values (i.e., >1s to avoid the race condition on
close that this timer addresses) are configured.
Risk Level: Medium
Testing: New unit tests added
Docs Changes: Updated version history and HttpConnectionManager proto doc
Fixes#6392
Signed-off-by: Andres Guedez <aguedez@google.com>
Mirrored from https://github.com/envoyproxy/envoy @ cdaeb1344e4136d8c9ac33507005159c93087a49
data-plane-api(CircleCI)