Skip to content

upstream: consistent H1 and H2 rq_total stats #6352

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
mattklein123 merged 2 commits into from
Mar 23, 2019
Merged

upstream: consistent H1 and H2 rq_total stats #6352

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
b97c5ed
upstream: consistent H1 and H2 rq_total stats
spenceral Mar 22, 2019
2976d0c
pr comment
spenceral Mar 22, 2019
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
8 changes: 6 additions & 2 deletions source/common/http/http1/conn_pool.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -90,9 +90,10 @@ void ConnPoolImpl::createNewConnection() {

ConnectionPool::Cancellable* ConnPoolImpl::newStream(StreamDecoder& response_decoder,
ConnectionPool::Callbacks& callbacks) {
host_->cluster().stats().upstream_rq_total_.inc();
host_->stats().rq_total_.inc();
if (!ready_clients_.empty()) {
host_->cluster().stats().upstream_rq_total_.inc();

@mattklein123 mattklein123 Mar 22, 2019

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.

Can you potentially move this into attachRequestToCleint() and remove from below? I think it's better to have the logic in one place and it's more accurate since we still may not send the request upstream if it gets cancelled while pending.

@spenceral spenceral Mar 22, 2019

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

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

This makes sense to me, thanks for catching that pending request case.

host_->stats().rq_total_.inc();

ready_clients_.front()->moveBetweenLists(ready_clients_, busy_clients_);
ENVOY_CONN_LOG(debug, "using existing connection", *busy_clients_.front()->codec_client_);
attachRequestToClient(*busy_clients_.front(), response_decoder, callbacks);
Expand All @@ -106,6 +107,9 @@ ConnectionPool::Cancellable* ConnPoolImpl::newStream(StreamDecoder& response_dec
host_->cluster().stats().upstream_cx_overflow_.inc();
}

host_->cluster().stats().upstream_rq_total_.inc();
host_->stats().rq_total_.inc();

// If we have no connections at all, make one no matter what so we don't starve.
if ((ready_clients_.size() == 0 && busy_clients_.size() == 0) || can_create_connection) {
createNewConnection();
Expand Down
3 changes: 3 additions & 0 deletions test/common/http/http1/conn_pool_test.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -293,6 +293,7 @@ TEST_F(Http1ConnPoolImplTest, MaxPendingRequests) {
cluster_->resetResourceManager(1, 1, 1024, 1);

EXPECT_EQ(0U, cluster_->circuit_breakers_stats_.rq_pending_open_.value());
EXPECT_EQ(0U, cluster_->stats_.upstream_rq_total_.value());

NiceMock<Http::MockStreamDecoder> outer_decoder;
ConnPoolCallbacks callbacks;
Expand All @@ -307,6 +308,7 @@ TEST_F(Http1ConnPoolImplTest, MaxPendingRequests) {
EXPECT_EQ(nullptr, handle2);

EXPECT_EQ(1U, cluster_->circuit_breakers_stats_.rq_pending_open_.value());
EXPECT_EQ(1U, cluster_->stats_.upstream_rq_total_.value());

handle->cancel();

Expand All @@ -315,6 +317,7 @@ TEST_F(Http1ConnPoolImplTest, MaxPendingRequests) {
dispatcher_.clearDeferredDeleteList();

EXPECT_EQ(1U, cluster_->stats_.upstream_rq_pending_overflow_.value());
EXPECT_EQ(1U, cluster_->stats_.upstream_rq_total_.value());
}

/**
Expand Down