Skip to content

Use local_info.node() instead of bootstrap.node() whenever possible #4120

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
htuch merged 7 commits into from
Aug 20, 2018
Merged

Use local_info.node() instead of bootstrap.node() whenever possible #4120

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
80b5498
Use local_info instead
dio Aug 13, 2018
6c263d3
Add tests
dio Aug 13, 2018
fd3071e
Use local_info for LoadStatsReporter
dio Aug 14, 2018
1792e22
Use local_info everywhere
dio Aug 14, 2018
0655cf8
Make sure timer is owned
dio Aug 14, 2018
53c3671
Merge remote-tracking branch 'upstream/master' into use-local-info
dio Aug 19, 2018
8b0dd55
review: change the thrown error message
dio Aug 19, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions source/common/upstream/cluster_manager_impl.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -211,7 +211,7 @@ ClusterManagerImpl::ClusterManagerImpl(const envoy::config::bootstrap::v2::Boots
// Now setup ADS if needed, this might rely on a primary cluster.
if (bootstrap.dynamic_resources().has_ads_config()) {
ads_mux_.reset(new Config::GrpcMuxImpl(
bootstrap.node(),
local_info.node(),

@ramaraochavali ramaraochavali Aug 13, 2018 •
edited
Loading

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can we add method similar to Config::Utility::checkLocalInfo("ads", local_info)
https://github.com/envoyproxy/envoy/blob/master/source/common/config/grpc_mux_impl.cc#L18
and validate that node has right info?

@dio dio Aug 13, 2018 •
edited
Loading

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, we can. However currently, if we failed to supply the info, it firstly throws an error for "cds", which I think serves the same purpose.

[2018-08-13 17:32:35.821][2350767][critical][main] source/server/server.cc:78] error initializing configuration '/envoy/config-local.yaml': cds: setting --service-cluster and --service-node is required

@dio dio Aug 14, 2018

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ramaraochavali after an experiment, I think it is tempting to have Config::Utility::checkLocalInfo inside the GrpcMuxImpl's constructor (and testing it is easier, yeay!). However, since we want to pass in the local_info here, the implication is quite large (but mostly "replacing" node with local_info). I have the changeset ready but want to gather opinions about it.

cc. @htuch @mattklein123.

Config::Utility::factoryForGrpcApiConfigSource(
*async_client_manager_, bootstrap.dynamic_resources().ads_config(), stats)
->create(),
Expand Down Expand Up @@ -301,7 +301,7 @@ ClusterManagerImpl::ClusterManagerImpl(const envoy::config::bootstrap::v2::Boots
if (cm_config.has_load_stats_config()) {
const auto& load_stats_config = cm_config.load_stats_config();
load_stats_reporter_.reset(
new LoadStatsReporter(bootstrap.node(), *this, stats,
new LoadStatsReporter(local_info.node(), *this, stats,

@htuch htuch Aug 14, 2018

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Another option, not saying you have to do this, is to treat the CLI flags as a bootstrap override and override the bootstrap proto early in server config, so that local_info and bootstrap are consistent. Or, maybe nuke node from bootstrap once you've loaded up local_info. Ideally we have a single source-of-truth, and it's hard to confuse with other potential sources.

@mattklein123 mattklein123 Aug 14, 2018

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

+1, it would be great to have a single source of truth here. From an interface perspective it seems better to have LocalInfo be the source of truth but I don't feel that strongly about it.

@dio dio Aug 15, 2018

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Updated to use local_info.

Config::Utility::factoryForGrpcApiConfigSource(
*async_client_manager_, load_stats_config, stats)
->create(),
Expand Down
4 changes: 4 additions & 0 deletions test/integration/ads_integration_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -132,6 +132,10 @@ class AdsIntegrationTest : public AdsIntegrationBaseTest,
envoy::api::v2::DiscoveryRequest discovery_request;
VERIFY_ASSERTION(ads_stream_->waitForGrpcMessage(*dispatcher_, discovery_request));

ASSERT(discovery_request.has_node());

@ramaraochavali ramaraochavali Aug 13, 2018 •
edited
Loading

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think while these ASSERTs are good, they would just validate existing tests right? Can we add a test in grpc_mux_impl_test that validates bad Node passed to it along the lines of this test

envoy/test/server/lds_api_test.cc

Line 203 in 0228180

TEST_F(LdsApiTest, BadLocalInfo) {

@dio dio Aug 13, 2018 •
edited
Loading

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks for this. Will try to read and add a test.

EXPECT_FALSE(discovery_request.node().id().empty());
EXPECT_FALSE(discovery_request.node().cluster().empty());

// TODO(PiotrSikora): Remove this hack once fixed internally.
if (!(expected_type_url == discovery_request.type_url())) {
return AssertionFailure() << fmt::format("type_url {} does not match expected {}",
Expand Down
3 changes: 3 additions & 0 deletions test/integration/load_stats_integration_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -164,6 +164,9 @@ class LoadStatsIntegrationTest : public HttpIntegrationTest,
return;
} else if (loadstats_request.cluster_stats_size() == 0) {
loadstats_request.CopyFrom(local_loadstats_request);
ASSERT(loadstats_request.has_node());

@htuch htuch Aug 14, 2018

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

You probably want ASSERT_TRUE to use the asserts from gtest.

EXPECT_FALSE(loadstats_request.node().id().empty());
EXPECT_FALSE(loadstats_request.node().cluster().empty());
return;
}

Expand Down