From 7917bbef78deb82b52e2cd86cdf6e141e5e21197 Mon Sep 17 00:00:00 2001
From: Eric Liang <ekhliang@gmail.com>
Date: Sun, 24 Nov 2019 22:37:59 -0800
Subject: [PATCH] Set progress report interval for bazel explicitly (#6262)

* set progress internval

* add keep alive

* add keepalive

* remove cat

* smaller time

* squash error

* reduce log spam
---
 .travis.yml                        |  2 +-
 ci/keep_alive                      | 30 ++++++++++++++++++++++++++++++
 src/ray/core_worker/core_worker.cc |  5 +++--
 3 files changed, 34 insertions(+), 3 deletions(-)
 create mode 100755 ci/keep_alive

diff --git a/.travis.yml b/.travis.yml
index 5531d53a7..afb2b3eeb 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -175,7 +175,7 @@ script:
   - if [ $RAY_CI_PYTHON_AFFECTED == "1" ]; then python -c 'import sys;exit(sys.version_info>=(3,5))' || python -m pytest -v --durations=5 --timeout=300 python/ray/tests/py3_test.py; fi
 
   # py bazel tests, run using local strategy since PY2 breaks with sandbox
-  - bazel test --spawn_strategy=local --python_version=$BAZEL_PYTHON_VERSION --incompatible_allow_python_version_transitions=false --incompatible_py3_is_default=false --show_progress_rate_limit=100 --show_timestamps --test_output=errors --test_tag_filters=-jenkins_only python/ray/...
+  - ./ci/keep_alive bazel test --spawn_strategy=local --python_version=$BAZEL_PYTHON_VERSION --incompatible_allow_python_version_transitions=false --incompatible_py3_is_default=false --progress_report_interval=100 --show_progress_rate_limit=100 --show_timestamps --test_output=errors --test_tag_filters=-jenkins_only python/ray/...
 
 deploy:
   - provider: s3
diff --git a/ci/keep_alive b/ci/keep_alive
new file mode 100755
index 000000000..5d86e4779
--- /dev/null
+++ b/ci/keep_alive
@@ -0,0 +1,30 @@
+#!/bin/bash
+# Run a command, printing periodically to keep travis alive.
+
+PID=$$
+
+# Print output to avoid travis killing us
+watchdog() {
+    for i in `seq 5 5 150`; do
+        sleep 300
+        echo "(running, ${i}m total)"
+    done
+    echo "Command timed out after 2.5h, dumping logs:"
+    echo "TIMED OUT"
+    kill -SIGKILL $PID
+}
+
+watchdog & 2>/dev/null
+WATCHDOG_PID=$!
+
+time "$@"
+
+CODE=$?
+if [ $CODE != 0 ]; then
+    echo "FAILED $CODE"
+    kill $WATCHDOG_PID
+    exit $CODE
+fi
+
+kill $WATCHDOG_PID
+exit 0
diff --git a/src/ray/core_worker/core_worker.cc b/src/ray/core_worker/core_worker.cc
index 5c6eba0fb..8c63e43cb 100644
--- a/src/ray/core_worker/core_worker.cc
+++ b/src/ray/core_worker/core_worker.cc
@@ -264,8 +264,9 @@ void CoreWorker::ReportActiveObjectIDs() {
   std::unordered_set<ObjectID> active_object_ids =
       reference_counter_->GetAllInScopeObjectIDs();
   RAY_LOG(DEBUG) << "Sending " << active_object_ids.size() << " object IDs to raylet.";
-  if (active_object_ids.size() > RayConfig::instance().raylet_max_active_object_ids()) {
-    RAY_LOG(WARNING) << active_object_ids.size() << " object IDs are currently in scope.";
+  auto max_active = RayConfig::instance().raylet_max_active_object_ids();
+  if (max_active && active_object_ids.size() > max_active) {
+    RAY_LOG(INFO) << active_object_ids.size() << " object IDs are currently in scope.";
   }
 
   if (!raylet_client_->ReportActiveObjectIDs(active_object_ids).ok()) {