posts/sampling-in-distributed-traces/ #14
Replies: 1 comment 1 reply
-
|
“因为需要大量内存来暂存数据,所以尾部采样对基础设施的要求很高,它减少了最终送达 Trace 后端落库的数据量,并且不像头部采样一样都是些没用的数据。” 这里说的数据是暂存在业务服务实例上还是监控系统服务上呢?如果是前者的话,压力应当比较分散,为什么说是基础设施要求高呢? |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
posts/sampling-in-distributed-traces/
https://jiekun.dev/posts/sampling-in-distributed-traces/
Beta Was this translation helpful? Give feedback.
All reactions