Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.
Original topic: ticdc链接kafka报错
Phenomenon:
When using TiCDC to synchronize data to downstream Kafka, the task runs normally. However, when a remote node is added to the downstream Kafka cluster, the TiCDC task starts intermittently reporting errors:
error=“[CDC:ErrKafkaNewSaramaProducer]new sarama producer: read tcp 10.108.182.13:46532->10.55.164.14:9092: i/o timeout”
After removing this remote node from Kafka, the task returns to normal and no longer reports errors. The network latency between the two locations is approximately 50ms.
This error is most likely due to network connectivity issues. Please check the network link issues from each TiCDC node to this new node.
At that time, I tested the network connectivity to 10.55 on the CDC node, and there was no problem.
You need to test the network from all CDC nodes to the target Kafka, as well as the network from the machine where the changefeed creation command is executed to Kafka.
All tests have been completed, no issues found.
What is this 10.108.182.13 address?
This is the address of the CDC node, which is also the address where TiUP is located.
I don’t know why either. According to the error message, it seems to be a network issue.
Check if there is a timeout parameter, you can try increasing it a bit.
Have you also confirmed that the firewall and ports are open?
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.
|
username: l940399478
|
username: 
|
username: 
|
username: 
|
username: 
|
username: