Defaulting Envoy mocks to StrictMocks. #1709
Conversation
…o today's tip-of-tree Signed-off-by: Alyssa Wilk <alyssar@chromium.org>
|
|
||
| set -e | ||
|
|
||
| VERSION=release-1.8.0 |
There was a problem hiding this comment.
I know that @jmillikin-stripe is doing work here, but can we stick with the existing prebuilt scripts for now? You can checkout the code directly from git and keep the rest of everything the same. See https://github.com/envoyproxy/envoy/blob/master/ci/build_container/build_recipes/boringssl.sh for an example of how this is done w/ Git. You will want to do a commit with just the switch to git/SHA for googletest. Then that will make a new build image. Then after that you can use that image SHA for this PR.
There was a problem hiding this comment.
You don't even need to switch to using Git checkouts. Change this single line to:
VERSION=43863938377a9ea1399c0596269e0890b5c5515a
There was a problem hiding this comment.
Neat trick I did not know that. Or that. :)
There was a problem hiding this comment.
oh, that would be fantastic as I think it would do minimal damage to our internal merges.
I'll send out a separate PR just for the hash change, then once docker updates I'll roll back the parts of this I can get rid of.
Signed-off-by: Alyssa Wilk <alyssar@chromium.org>
…xy#1715) Signed-off-by: Alyssa Wilk <alyssar@chromium.org>
| ":" + name + "_lib", | ||
| repository + "//test:main" | ||
| ], | ||
| args = args + ["--gmock_default_mock_behavior=2"], |
There was a problem hiding this comment.
It's hard to find the meaning of "2" here, because "gmock_default_mock_behavior" doesn't appear anywhere in the googletest codebase (it's defined with a special macro). Can you copy the meanings from that flag's --help into here as a comment?
Signed-off-by: Alyssa Wilk <alyssar@chromium.org>
Description: Add linux config for Engflow's remote execution. No semantic changes. This PR is a no-op. We add two things to the envoy-mobile build: - A Linux-specific bazelrc configuration that sets all the necessary flags to run on a Linux Engflow worker - Instruct envoy to use the latest envoyproxy/envoy-build-ubuntu docker container for the linux builds through the WORKSPACE rules Notice that the envoy image is only necessary for the Linux builds since the MacOS builds are set somehow manually and don't use containers. Also notice that Engflow builds have a notion of "Pool" which for now only apply to Linux workers since the first workers in the Envoy cluster were Macs and they didn't have a Pool associated to it (e.g. no pool means Mac at the moment). Once the MacOS workers belong to a Pool we'll update the MacOS config. Risk Level: Low Testing: No-op Docs Changes: N/A Release Notes: N/A Signed-off-by: Luis Fernando Pino Duque <luis@engflow.com> Signed-off-by: JP Simard <jp@jpsim.com>
Description: Add linux config for Engflow's remote execution. No semantic changes. This PR is a no-op. We add two things to the envoy-mobile build: - A Linux-specific bazelrc configuration that sets all the necessary flags to run on a Linux Engflow worker - Instruct envoy to use the latest envoyproxy/envoy-build-ubuntu docker container for the linux builds through the WORKSPACE rules Notice that the envoy image is only necessary for the Linux builds since the MacOS builds are set somehow manually and don't use containers. Also notice that Engflow builds have a notion of "Pool" which for now only apply to Linux workers since the first workers in the Envoy cluster were Macs and they didn't have a Pool associated to it (e.g. no pool means Mac at the moment). Once the MacOS workers belong to a Pool we'll update the MacOS config. Risk Level: Low Testing: No-op Docs Changes: N/A Release Notes: N/A Signed-off-by: Luis Fernando Pino Duque <luis@engflow.com> Signed-off-by: JP Simard <jp@jpsim.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com> Signed-off-by: Ryan Northey <ryan@synca.io>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com> Signed-off-by: Leonardo da Mata <ldamata@spotify.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com>
Fix for GHSA-g9vw-6pvx-7gmw Signed-off-by: Yan Avlasov <yavlasov@google.com> Signed-off-by: Misha Badov <mbadov@google.com>
**Description** Add policy for a backed token quota. Configuring multiple backends in an AIGatewayRouteRule allows request that exceeds allotted token quota on one service to fall back to the next service in the list. **Related Issues/PRs (if applicable)** Part of: #1571 **Special notes for reviewers (if applicable)** This PR contains just API change to make sure this is the right direction. I will add tests once we agree on the API. Split configuration between QutaPolicy with the metadata keys and additional BackendTrafficPolicy is cumbersome. Perhaps BackendTrafficPolicy content should just be made part of the QutaPolicySpec for simplicity. --------- Signed-off-by: Yan Avlasov <yavlasov@google.com>
[googletest updated to tip-of-tree in #1715]
Defaulting envoy tests to use StrictMock semantics by default
Fixing up all the mock tests (since they now fail)
Fixes #1695