-
Notifications
You must be signed in to change notification settings - Fork 2.2k
Description
Search before asking
- I had searched in the issues and found no similar issues.
What happened
During my use of the mysql-cdc, I discovered a phenomenon. Firstly, I need to explain my environment. My source and target are both MySQL, and I have a batch of inventory data. In the process of fully synchronizing inventory data, there will be no problem if no new data is added to the source database. But if during this period, I have a batch of new data that needs to be inserted into the source library, during which time the JVM memory and MySQL connections will continue to grow. When this process is long enough, it can cause JVM memory to approach the threshold, triggering GC, but GC cannot effectively release memory. At the same time, it will cause excessive consumption of MySql Connection pool.
SeaTunnel Version
seatunnel-2.3.2
SeaTunnel Config
env {
# You can set SeaTunnel environment configuration here
job.name = "mysql_test"
job.mode = "STREAMING"
checkpoint.interval = 10000
execution.checkpoint.interval = 10000
execution.checkpoint.data-uri = "hdfs://localhost:9000/checkpoint"
}
source {
MySQL-CDC {
result_table_name = "mysql_cdc_test"
snapshot.split.size = 3000
incremental.parallelism = 1
server-id = "5400"
username = "xxxx"
password = "xxxxxxxx"
database-names = ["xxxxx"]
table-names = ["xxxxx.test_data"]
base-url = "jdbc:mysql://192.168.xxx.xxx:xxxx/xxxxx"
}
}
sink{
jdbc {
url = "jdbc:mysql://192.168.xxx.xxx:xxxx/xxxxx"
driver = "com.mysql.cj.jdbc.Driver"
user = "xxxx"
password = "xxxxxx"
table = "xxxxx"
primary_keys = ["id"]
database = "xxxxxxxx"
batch_size = 3000
batch_interval_ms = 20
}
}
Running Command
./bin/seatunnel.sh -c config/v2.mysql_cdc.configError Exception
No obvious abnormality
Flink or Spark Version
No response
Java or Scala Version
Jdk11
Screenshots
No response
Are you willing to submit PR?
- Yes I am willing to submit a PR!
Code of Conduct
- I agree to follow this project's Code of Conduct