[jira] [Created] (IOTDB-5248) [ratis] Restarting datanode takes a long time (reading raft log)

["刘珍 (Jira)" <ji...@apache.org>]

刘珍 created IOTDB-5248:
-------------------------

             Summary: [ratis] Restarting datanode takes a long time (reading raft log)
                 Key: IOTDB-5248
                 URL: https://issues.apache.org/jira/browse/IOTDB-5248
             Project: Apache IoTDB
          Issue Type: Bug
          Components: mpp-cluster
    Affects Versions: master branch
            Reporter: 刘珍
            Assignee: Xinyu Tan


rel/1.0 1220 00a4080
1. 启动3副本3C5D集群，config/schema/data均是ratis协议
2.BM写入，完成。
3.ip73 cli flush
stop-datanode.sh  take snapshot成功。
start-datanode.sh 启动耗时过长，查看log，一直刷 读raft log的日志（stop datanode，打的快照不起作用？）
 !image-2022-12-20-15-08-34-025.png! 
测试环境
1. 192.168.10.62/66/68 3ConfigNode 72cpu 256GB
192.168.10.62/66/68/64/73 5DataNode
73机器：48CPU 384GB

2.数据库配置参数
COMMON配置
schema_replication_factor=3
data_replication_factor=3
data_region_consensus_protocol_class=org.apache.iotdb.consensus.ratis.RatisConsensus
query_timeout_threshold=3600000
ConfigNode配置
cn_connection_timeout_ms=120000
MAX_HEAP_SIZE="8G"

DataNode配置
MAX_HEAP_SIZE="192G"
MAX_DIRECT_MEMORY_SIZE="32G"
dn_max_connection_for_internal_service=300
3.BM配置见附件
写入完成
4.ip73
stop-datanode.sh
清缓存，启动datanode。




--
This message was sent by Atlassian Jira
(v8.20.10#820010)