root
/
kafka
mirror of https://github.com/apache/kafka.git
1
0
Fork 0
Code Issues Actions 2 Packages Projects Releases Wiki Activity

MINOR: Recommend Java 11 (#9080) 

Java 11 has been recommended for a while, but the ops section had not been updated.
Also added `-XX:+ExplicitGCInvokesConcurrent` which has been in `kafka-run-class`
for a while. Finally, tweaked the text slightly to read better.

Reviewers: Manikumar Reddy <manikumar.reddy@gmail.com>
This commit is contained in:
Ismael Juma 2020-07-26 12:25:30 -07:00 committed by GitHub
parent 0d47c69a93
commit 6004ee9f22
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 9 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
docs/ops.html
View File

@ -680,16 +680,20 @@
<h3><a id="java" href="#java">6.4 Java Version</a></h3>
From a security perspective, we recommend you use the latest released version of JDK 1.8 as older freely available versions have disclosed security vulnerabilities.
Java 8 and Java 11 are supported. Java 11 performs significantly better if TLS is enabled, so it is highly recommended (it also includes a number of other
performance improvements: G1GC, CRC32C, Compact Strings, Thread-Local Handshakes and more).
From a security perspective, we recommend the latest released patch version as older freely available versions have disclosed security vulnerabilities.
At the time this is written, LinkedIn is running JDK 1.8 u5 (looking to upgrade to a newer version) with the G1 collector. LinkedIn's tuning looks like this:
Typical arguments for running Kafka with OpenJDK-based Java implementations (including Oracle JDK) are:
<pre class="brush: text;">
-Xmx6g -Xms6g -XX:MetaspaceSize=96m -XX:+UseG1GC
-XX:MaxGCPauseMillis=20 -XX:InitiatingHeapOccupancyPercent=35 -XX:G1HeapRegionSize=16M
-XX:MinMetaspaceFreeRatio=50 -XX:MaxMetaspaceFreeRatio=80
-XX:MinMetaspaceFreeRatio=50 -XX:MaxMetaspaceFreeRatio=80 -XX:+ExplicitGCInvokesConcurrent
</pre>
For reference, here are the stats on one of LinkedIn's busiest clusters (at peak):
For reference, here are the stats for one of LinkedIn's busiest clusters (at peak) that uses said Java arguments:
<ul>
<li>60 brokers</li>
<li>50k partitions (replication factor 2)</li>
@ -697,7 +701,7 @@
<li>300 MB/sec inbound, 1 GB/sec+ outbound</li>
</ul>
The tuning looks fairly aggressive, but all of the brokers in that cluster have a 90% GC pause time of about 21ms, and they're doing less than 1 young GC per second.
All of the brokers in that cluster have a 90% GC pause time of about 21ms with less than 1 young GC per second.
<h3><a id="hwandos" href="#hwandos">6.5 Hardware and OS</a></h3>
We are using dual quad-core Intel Xeon machines with 24GB of memory.