Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.
Original topic: drainer进程在,端口无法启动
[TiDB Usage Environment] Production Environment
[TiDB Version]
Upstream TiDB Version: v3.0.3
Downstream TiDB Version: v5.2.4
[Overview] Scenario + Problem Overview
Incremental data from the upstream TiDB is synchronized to the downstream higher version cluster through the drainer service. The drainer service is deployed via ansible, but it fails to start. Checking the process with ps shows it is running, but the port is not opened.
[Resource Configuration]
Drainer server: 203G memory; 10Gbps network interface; 32 cores;
All pump statuses are online:
Binlog status is enabled:
Memory is not fully utilized:
Bandwidth situation:
It looks fine. Suggest adjusting the wait time and restarting.
Wait a bit, when drainer starts, it will traverse the entire DDL history. If there are many DDLs in this cluster… it might take a very long time.
However, this ansible-playbook will report an error and stop.
As this brother said, because Ansible waits for a period of time by default, if it doesn’t start successfully within that time, it will report an error and exit. In reality, it is still in the process of starting up.
Okay, then I will continue to monitor the drainer.log log.
No, after the memory usage goes up, OOM kills the drainer.
Upgrade the memory. In cases with many DDL operations, a lot of memory is needed.
Okay, I’ll think about how to do it.
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.