刘珍 created IOTDB-4752:
-------------------------

             Summary: some data is lost after the read-only node returns to 
normal
                 Key: IOTDB-4752
                 URL: https://issues.apache.org/jira/browse/IOTDB-4752
             Project: Apache IoTDB
          Issue Type: Bug
          Components: mpp-cluster
    Affects Versions: 0.14.0-SNAPSHOT
            Reporter: 刘珍
            Assignee: 张洪胤
         Attachments: image-2022-10-25-16-22-00-198.png

m_1025_cbc6225
schema region : ratis
data region : multiLeader
3rep , 3C3D
During benchmark writing, ip72 is set to read-only due to "no space left on 
device". 
Wait for the benchmark write to complete, ip72 release disk space.
ip72 : SET SYSTEM TO RUNNING
Wait for synchronization to complete
ip72 : flush
Perform query comparison,ip72 lost some data.
"select count(s_0) ,count(s_9),count(s_99),count(s_999),count(s_9999) from 
root.** align by device"
 !image-2022-10-25-16-22-00-198.png! 

Test environment :
1. 192.168.10.72 / 73 /74  48CPU 384GB
benchmark : ip75 /home/liuzhen/benchmark/bm_0620_7ec96c1
iotdb_dir : /ssd_data/mpp_test/m_1025_cbc6225

ConfigNode
MAX_HEAP_SIZE="8G"
schema_region_consensus_protocol_class=org.apache.iotdb.consensus.ratis.RatisConsensus
data_region_consensus_protocol_class=org.apache.iotdb.consensus.multileader.MultiLeaderConsensus
schema_replication_factor=3
data_replication_factor=3
connection_timeout_ms=120000
schema_region_ratis_request_timeout_ms=1200000
data_region_ratis_request_timeout_ms=1200000
schema_region_ratis_max_retry_attempts=1
data_region_ratis_max_retry_attempts=2

DataNode
MAX_HEAP_SIZE="256G"
MAX_DIRECT_MEMORY_SIZE="32G"
avg_series_point_number_threshold=1
max_waiting_time_when_insert_blocked=3600000
enable_seq_space_compaction=false
enable_unseq_space_compaction=false
enable_cross_space_compaction=false
query_timeout_threshold=36000000


2. benchmark configuration
see attachment .

3. During benchmark writing, ip72 is set to read-only due to "no space left on 
device". 

4. Wait for the benchmark write to complete, ip72 release disk space.

5.  ip72 : SET SYSTEM TO RUNNING

6.  Wait for synchronization to complete

7.  ip72 : flush

8. Perform query comparison
select count(s_0) ,count(s_9),count(s_99),count(s_999),count(s_9999) from 
root.** align by device



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

Reply via email to