Aggarwal-Raghav commented on code in PR #5884:
URL: https://github.com/apache/hive/pull/5884#discussion_r2160237406
##########
conf/hive-env.sh.template:
##########
@@ -18,18 +20,29 @@
# to control the execution of Hive. It should be used by admins to configure
# the Hive installation (so that users do not have to set environment variables
# or set command line parameters to get correct behavior).
-#
-# The hive service being invoked (CLI etc.) is available via the environment
-# variable SERVICE
+# Set HADOOP_HOME to point to a specific hadoop install directory
+# HADOOP_HOME=${bin}/../../hadoop
+
+# Hive Configuration Directory can be controlled by:
+# export HIVE_CONF_DIR=
+
+# Folder containing extra libraries required for hive compilation/execution
can be controlled by:
+# export HIVE_AUX_JARS_PATH=
+
+# HIVE Log Directory can be controlled by and ensure it exists on the
filesystem:
+# export HIVE_LOG_DIR=
+
+# The hive service being invoked (beeline etc.) is available via the
environment
+# variable SERVICE
# Hive Client memory usage can be an issue if a large number of clients
-# are running at the same time. The flags below have been useful in
+# are running at the same time. The flags below have been useful in
# reducing memory usage:
-#
-# if [ "$SERVICE" = "cli" ]; then
+
+# if [ "$SERVICE" = "beeline" ]; then
# if [ -z "$DEBUG" ]; then
-# export HADOOP_OPTS="$HADOOP_OPTS -XX:NewRatio=12 -Xms10m
-XX:MaxHeapFreeRatio=40 -XX:MinHeapFreeRatio=15 -XX:+UseParNewGC
-XX:-UseGCOverheadLimit"
+# export HADOOP_OPTS="$HADOOP_OPTS -XX:NewRatio=12 -Xms10m
-XX:MaxHeapFreeRatio=40 -XX:MinHeapFreeRatio=15 -XX:+G1GC
-XX:-UseGCOverheadLimit"
Review Comment:
Thanks for the review @okumin. Few points:
1. Yes -`XX:+UseG1GC` is the correct one. I doubled confirmed using:
```java -XX:+UnlockExperimentalVMOptions -XX:+PrintFlagsFinal -version |
grep -i "G1GC"```
2. Regarding removing the GC flags and NewRatio, I am inclined towards
removing every flag other than Xmx and Xms. Reason being, G1GC and ZGC doesn't
require much tuning and tuning should be system specific + workload specifc
which varies cluster to cluster.
3. I "want to believe" that the DEBUG option is there to showcase user the
possibilities / kind of tuning users can do.
Let me know your thoughts based on that I will update the PR.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: gitbox-unsubscr...@hive.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: gitbox-unsubscr...@hive.apache.org
For additional commands, e-mail: gitbox-h...@hive.apache.org
