Grafana 展示cpu 显示nodata

[TiDBer_yyy]

监控路径: overview–> tikv → cpu
版本：5.0.4
安装方法：tiup安装
问题：查看各个instance的cpu适用情况，发现nodata，并提示：

"query processing would load too many samples into memory in query execution"

[songxuecheng (小白)]

tiup cluster reload xxx -R prometheus 试试 并且检查一下监控机器负载情况

[TiDBer_yyy]

重启后未生效。

监控机器负载较低

PS： 使用环境中将 IP 替换为 hostname, 其他监控指标能显示对应的hostname。

sed -i.bak -e "s/172.0.0.1\b/hostname1/g" -e "s/172.0.0.2\b/hostname2/g" prometheus.yml

[songxuecheng (小白)]

prometheus的日志发一下

[TiDBer_yyy]

tmp_prom.log (1.4 MB)

[Aric (Jansu Dev)]

@TiDBer_yangky

query processing would load too many samples into memory in query execution

Prometheus PR : Getting "query processing would load too many samples" error for simple count(metric) query · Issue #7281 · prometheus/prometheus · GitHub
规避方法研究 ： prometheus 查询30d数据出现query processing would load too many samples问题_query processing would load too many samples into -CSDN博客

是改过 TiDB 官方默认的 Prometheus 配置吗？

[TiDBer_yyy]

感谢，问题已解决