[bookkeeper] 13/31: Enable PCBC completionObjects autoShrink to reduce memory usage and gc (#3913)

[yo...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

yong pushed a commit to branch branch-4.16
in repository https://gitbox.apache.org/repos/asf/bookkeeper.git

commit 8c4f3c8cbb36cc8fe272dfc2916b6fa7cce98ffe
Author: wenbingshen <ol...@gmail.com>
AuthorDate: Sat Apr 22 10:23:52 2023 +0800

    Enable PCBC completionObjects autoShrink to reduce memory usage and gc (#3913)
    
    ### Motivation
    
    PerChannelBookieClient completionObjects occupy a lot of heap space and cannot be recycled.
    The figure below shows that the internal table array of ConcurrentOpenHashMap has used space size=0, but the array length is still 16384, and the memory overhead is 65552bytes.
    ![image](https://user-images.githubusercontent.com/35599757/231114802-db90c49b-d295-46d7-b7db-785035b341f0.png)
    
    ![image](https://user-images.githubusercontent.com/35599757/231113930-bd9f3f54-9052-4c0b-9a3f-2fc493632e35.png)
    
    ConcurrentOpenHashMap default DefaultConcurrencyLevel=16. We have hundreds of bookie nodes. Due to the feature of bookie polling and writing, the client and server have long connection characteristics, which will as a result, the memory usage of about 65552 * 16 * 1776 = 1.74GB cannot be recycled, and the space take up by these tables is all size=0 (The broker's owner topic has drifted to other brokers due to Full GC).
    ![image](https://user-images.githubusercontent.com/35599757/231117087-08c80320-fa71-49c2-a199-cfee3d83ddc5.png)
    
    When the throughput of the pulsar cluster increases and the bookie cluster expands, these memory usage will also increase. Coupled with the unreasonable memory usage in other aspects of pulsar that we know, this will cause the pulsar broker to continuously generate full gc.
    
    ### Changes
    I think adding autoShrink to completionObjects can reduce this part of memory usage and reduce the frequency of Full GC.
    
    (cherry picked from commit ca33b31d3918d04bee63fba54b81b531709a6536)
---
 .../main/java/org/apache/bookkeeper/proto/PerChannelBookieClient.java   | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/bookkeeper-server/src/main/java/org/apache/bookkeeper/proto/PerChannelBookieClient.java b/bookkeeper-server/src/main/java/org/apache/bookkeeper/proto/PerChannelBookieClient.java
index 41add30ca5..f84343a71b 100644
--- a/bookkeeper-server/src/main/java/org/apache/bookkeeper/proto/PerChannelBookieClient.java
+++ b/bookkeeper-server/src/main/java/org/apache/bookkeeper/proto/PerChannelBookieClient.java
@@ -187,7 +187,7 @@ public class PerChannelBookieClient extends ChannelInboundHandlerAdapter {
     final int startTLSTimeout;
 
     private final ConcurrentOpenHashMap<CompletionKey, CompletionValue> completionObjects =
-            ConcurrentOpenHashMap.<CompletionKey, CompletionValue>newBuilder().build();
+            ConcurrentOpenHashMap.<CompletionKey, CompletionValue>newBuilder().autoShrink(true).build();
 
     // Map that hold duplicated read requests. The idea is to only use this map (synchronized) when there is a duplicate
     // read request for the same ledgerId/entryId