Note:
This topic has been translated from a Chinese forum by GPT and might contain errors.Original topic: Trino查询Tidb出现以下错误
|
username: TIDB救我狗命
[TiDB Usage Environment] Production Environment / Testing / PoC
Production Environment
[TiDB Version]
6.0.1
[Reproduction Path] What operations were performed when the issue occurred
Trino querying TiDB
[Encountered Issue: Issue Phenomenon and Impact]
Currently, I can’t see the cluster status. I might need to wait for the operations team to check it. A few days ago, we upgraded the TiDB cluster, and now the situation is that some tables can be queried, while others are experiencing the aforementioned issue.
Suspect that TiKV has crashed and there is data corruption. Need to check the specific cluster status.
Yesterday, the issue was queried at 00:16, and then TiKV restarted once at 00:31. This morning, I encountered a query failure at 09:12, but between 09:12 and 09:27, some queries succeeded while others failed. Then, TiKV restarted once at 09:27. The timing doesn’t seem to match up.
Take a detailed look at the logs of TiDB and TiKV to determine the cause of the failure. When recovering, back up the data and try to scale down this KV node and then scale up a new one.
I asked the operations team, and the restart was due to an OOM (Out of Memory) event.
The operations team said it seems like the page size is insufficient.
You have quite a few regions. Finally, adjust the heartbeat and enable region merge and cross-table region merge.
If you find that the merge is not happening, you need to increase the scheduling-related and merge-related parameters.
Oh, okay, but it seems that the issue I’m encountering is that the leader cannot be found… So strange.
The image you provided is not visible. Please provide the text you need translated.
It might be due to scheduling issues. Follow my suggestion to reduce the pressure on TiKV first; excessive heartbeats are a problem.
|
username: 
|
username: 