From 4cfe1467011dfd15dd781c981decf31e109fdede Mon Sep 17 00:00:00 2001
From: Marcelo Zimbres <mzimbres@gmail.com>
Date: Sun, 18 Aug 2024 11:11:05 +0200
Subject: [PATCH] Improves reconnection responsiveness.

As the user points out in

   https://github.com/boostorg/redis/issues/205

wait_for_all might be the cause of delay in the
reconnection. I could not observe any improvement in my
local setup but wait_for_one_error look the correct token
and all tests pass so I am willing to change it.
---
 README.md                                      | 6 ++++++
 include/boost/redis/detail/connection_base.hpp | 2 +-
 include/boost/redis/detail/runner.hpp          | 2 +-
 test/test_conn_exec.cpp                        | 3 +++
 4 files changed, 11 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index 8634425d..7f500c79 100644
--- a/README.md
+++ b/README.md
@@ -676,6 +676,12 @@ https://lists.boost.org/Archives/boost/2023/01/253944.php.
 
 ## Changelog
 
+### Boost 1.87
+
+* ([Issue 205](https://github.com/boostorg/redis/issues/205))
+  Improves reaction time to disconnection by using `wait_for_one_error`
+  instead of `wait_for_all`.
+
 ### Boost 1.85
 
 * ([Issue 170](https://github.com/boostorg/redis/issues/170))
diff --git a/include/boost/redis/detail/connection_base.hpp b/include/boost/redis/detail/connection_base.hpp
index 487c759f..e72c1f16 100644
--- a/include/boost/redis/detail/connection_base.hpp
+++ b/include/boost/redis/detail/connection_base.hpp
@@ -317,7 +317,7 @@ struct reader_op {
             if (ec == asio::error::eof) {
                logger_.trace("reader-op: EOF received. Exiting ...");
                conn_->cancel(operation::run);
-               return self.complete({}); // EOFINAE: EOF is not an error.
+               return self.complete(ec);
             }
 
             // The connection is not viable after an error.
diff --git a/include/boost/redis/detail/runner.hpp b/include/boost/redis/detail/runner.hpp
index f257019b..de185e71 100644
--- a/include/boost/redis/detail/runner.hpp
+++ b/include/boost/redis/detail/runner.hpp
@@ -93,7 +93,7 @@ class runner_op {
             [this](auto token) { return runner_->health_checker_.async_check_health(*conn_, logger_, token); },
             [this](auto token) { return runner_->async_hello(*conn_, logger_, token); }
          ).async_wait(
-            asio::experimental::wait_for_all(),
+            asio::experimental::wait_for_one_error(),
             std::move(self));
 
          logger_.on_runner(ec0, ec1, ec2);
diff --git a/test/test_conn_exec.cpp b/test/test_conn_exec.cpp
index c3f04134..e55c325a 100644
--- a/test/test_conn_exec.cpp
+++ b/test/test_conn_exec.cpp
@@ -152,6 +152,9 @@ BOOST_AUTO_TEST_CASE(correct_database)
    auto const pos = value.find("db=");
    auto const index_str = value.substr(pos + 3, 1);
    auto const index = std::stoi(index_str);
+
+   // This check might fail if more than one client is connected to
+   // redis when the CLIENT LIST command is run. 
    BOOST_CHECK_EQUAL(cfg.database_index.value(), index);
 }