The effective time of space is not friendly to testing
jxnu-liguobin opened this issue · 2 comments
Is your feature request related to a problem? Please describe.
I am testing the Nebula application by testcontainers. Of course, testcontainers do not support nebula now, I have written a simple tool testcontainers-nebula.
The document suggests that creating a space requires a wait of 20 seconds before it becomes available, which is good for distributed systems, but it is cumbersome for testing, which means that you must retry or manually sleep. The latter is not easily achieved correctly in asynchronous programs, and the number of retries for the former is unknown.
For this issue, in zio, I had to use the eventually aspect to make my testing ultimately successful (Try again until successful), obviously, this is time-consuming and unknown, and may never succeed.
heartbeat_interval_secs It seems that the configuration controls this, but can it take effect immediately?
Describe the solution you'd like
For this, allowing clients to register callbacks to inform users in a timely manner or effective immediately?.
Describe alternatives you've considered
Additional context
heartbeat_interval_secs It seems that the configuration controls this, but can it take effect immediately?
The configuration is static, must restart graphd server to make it take effect.
@jxnu-liguobin Thanks for your test tool, It will be a friendly and effective tool for users to make nebula related test. Could you please contribute your great test container to nebula-contrib? cc @QingZ11 @wey-gu
Wow, second project in a role!!!!🚀