From 50dbf1a3078b953ebaaad4ed9432a9302df2757b Mon Sep 17 00:00:00 2001
From: Richard Liaw <rliaw@berkeley.edu>
Date: Tue, 10 Nov 2020 14:57:57 -0800
Subject: [PATCH] [core] Support configurable number of "check for redis"
 attempts (#11902)

Signed-off-by: Richard Liaw <rliaw@berkeley.edu>
---
 python/ray/_private/services.py | 12 ++++++++----
 python/ray/ray_constants.py     |  4 ++++
 2 files changed, 12 insertions(+), 4 deletions(-)

diff --git a/python/ray/_private/services.py b/python/ray/_private/services.py
index 13116dff2..fa0545a2e 100644
--- a/python/ray/_private/services.py
+++ b/python/ray/_private/services.py
@@ -551,7 +551,7 @@ def wait_for_redis_to_start(redis_ip_address, redis_port, password=None):
     redis_client = redis.StrictRedis(
         host=redis_ip_address, port=redis_port, password=password)
     # Wait for the Redis server to start.
-    num_retries = 12
+    num_retries = ray_constants.START_REDIS_WAIT_RETRIES
     delay = 0.001
     for _ in range(num_retries):
         try:
@@ -567,9 +567,13 @@ def wait_for_redis_to_start(redis_ip_address, redis_port, password=None):
         else:
             break
     else:
-        raise RuntimeError("Unable to connect to Redis. If the Redis instance "
-                           "is on a different machine, check that your "
-                           "firewall is configured properly.")
+        raise RuntimeError(
+            f"Unable to connect to Redis (after {num_retries} retries). "
+            "If the Redis instance is on a different machine, check that "
+            "your firewall and relevant Ray ports are configured properly. "
+            "You can also set the environment variable "
+            "`RAY_START_REDIS_WAIT_RETRIES` to increase the number of "
+            "attempts to ping the Redis server.")
 
 
 def _compute_version_info():
diff --git a/python/ray/ray_constants.py b/python/ray/ray_constants.py
index 29f5387b5..c748d4b4f 100644
--- a/python/ray/ray_constants.py
+++ b/python/ray/ray_constants.py
@@ -137,6 +137,10 @@ RESOURCES_ENVIRONMENT_VARIABLE = "RAY_OVERRIDE_RESOURCES"
 # The reporter will report its statistics this often (milliseconds).
 REPORTER_UPDATE_INTERVAL_MS = env_integer("REPORTER_UPDATE_INTERVAL_MS", 2500)
 
+# Number of attempts to ping the Redis server. See
+# `services.py:wait_for_redis_to_start`.
+START_REDIS_WAIT_RETRIES = env_integer("RAY_START_REDIS_WAIT_RETRIES", 12)
+
 LOGGER_FORMAT = (
     "%(asctime)s\t%(levelname)s %(filename)s:%(lineno)s -- %(message)s")
 LOGGER_FORMAT_HELP = f"The logging format. default='{LOGGER_FORMAT}'"