roger
roger
@CastroKKK 要是有 graph 挂了，挂了的 graph 节点数据是不是都会丢了。
@CastroKKK 多谢。大致看了下，就是做个镜像(副本)了，如果发现挂了的，得手动踢掉换上镜像，这样流量也翻倍了，如果是To C，感觉这个恢复手法有点不能接受呐。感觉还是需要一个完整的分布式时序数据库呐。
@beorn7
@Joffcom fullRunData is like this. ``` { "data": { "startData": {}, "resultData": { "runData": { "Schedule Trigger": [{ "startTime": 1706711194187, "executionTime": 2, "source": [], "executionStatus": "success", "data": { "main": [...
@roidelapluie I manually exposed the data with a timestamp. My original data is inconsistent with the data queried by prometheus. Is this okay? I don't quite understand what you mean.
@roidelapluie Isn't this it?I don't know what you're going to say. [[1721872802,"14"], [1721872862,"14"], [1721872922,"19"], [1721872982,"15"], [1721873042,"15"], [1721873102,"13"], [1721873162,"11"], [1721873222,"11"], [1721873282,"18"], [1721873342,"15"], [1721873402,"15"]]
> How do you send those metrics to Prometheus? At once? @roidelapluie yes