@wu-sheng 

I mean its cost is too expensive, about k8s and docker, we are just try to use, 
they are only used in the test env. Just have a try, we are not very familiar 
with them, I just learn to use them.

And all test app are deployed in the k8s, if I restore to the original crime 
scene, it will cost too time. Because the k8s are used and dockers are deployed 
every day, its hard to restore to original scene.

And another reason is that it will have a big influence to our business 
development progress, and cost several colleague time, at least one day.The 
worst is that maybe still have no chance to restore it.
Because I guess:
This problem ought to occur in some extreme case: when io, mem resources 
arrived at some critical point.
It becomes much more difficult to restore it.

And in the other word, In china, especially for not big startup company, its so 
expensive. 
Especially we almost confirm where it is and delta resources are limited in a 
constant range.

So I think we could confirm effect with gperftools analysis result instead.I 
think this is the efficiency way to solve the problem.

[ Full content available at: 
https://github.com/apache/incubator-skywalking/issues/1666 ]
This message was relayed via gitbox.apache.org for [email protected]

Reply via email to