These are chat archives for spring-cloud/spring-cloud

10th
May 2017
Miguel Angel Chico
@miguelchico
May 10 2017 13:15

Hi. I'm updating from spring-cloud Camden.SR5 to Dalston and I have a problem that can't understand.
Just changing the version and trying to update I get this error:

***************************
APPLICATION FAILED TO START
***************************

Description:

Parameter 0 of method eurekaHealthIndicator in org.springframework.cloud.netflix.eureka.EurekaDiscoveryClientConfiguration$EurekaHealthIndicatorConfiguration required a bean of type 'com.netflix.discovery.EurekaClient' that could not be found.
    - Bean method 'eurekaClient' not loaded because Ancestor org.springframework.cloud.netflix.eureka.EurekaClientAutoConfiguration did not match


Action:

Consider revisiting the conditions above or defining a bean of type 'com.netflix.discovery.EurekaClient' in your configuration.

I had the annotation @EnableDiscoveryClient in an autoconfiguration class as part as a dependency. The configuration is loaded but raise the error. If I move the annotation from the dependency to my Main class then the application loads.
Do you have any idea?

Stian Lågstad
@stianlagstad
May 10 2017 14:12
Could someone assist me in getting custom hystrix timeouts to work? I've created an example here: https://github.com/stianlagstad/spring-boot-timeout-demo . I want to set a low timeout for a feign client in a test, but I'm obviously not understanding how I should do is because this doesn't work: https://github.com/stianlagstad/spring-boot-timeout-demo/blob/master/src/test/java/com/example/TimeoutDemoApplicationTests.java#L30
Dave Syer
@dsyer
May 10 2017 14:27
@miguelchico if you put conditions on autoconfig classes you have to add an explicit order
In this case you need it to run before something in eureka client config
Or after.
After probably.
(since userspace configs are processed last)
Miguel Angel Chico
@miguelchico
May 10 2017 14:45
Thanks @dsyer , that's what I thought, the thing is I'm trying to figure out the right order. Thank you
Nastya Smirnova
@Aloren
May 10 2017 16:01
Hi all! Found an issue with eureka health indicator. It says "failCount": 374364. Looked into the code, but I can't understand what it does: https://github.com/spring-cloud/spring-cloud-netflix/blob/master/spring-cloud-netflix-eureka-client/src/main/java/org/springframework/cloud/netflix/eureka/EurekaHealthIndicator.java#L80 Could please someone explain? Does that mean that service fetches eureka registry too often? Thanks in advance!
Dave Syer
@dsyer
May 10 2017 16:08
No, it means that the app hasn't been able to fetch an updated catalog from the eureka server (for more than the threshold time specified in registry fetch interval).
The failCountdoesn't look very helpful to me
The larger it is the worse the problem
But the absolute magnitude isn't very helpful
and it looks like there is a millisecond/second factor that inflates it as well
Nastya Smirnova
@Aloren
May 10 2017 16:12
Hm
I am able to see refresh events from this application on Eureka server
And there are no exceptions in logs
Dave Syer
@dsyer
May 10 2017 16:12
The health indicator is UP, right?
Nastya Smirnova
@Aloren
May 10 2017 16:13
yes. BTW DiscoveryClientHealthIndicator is reporting correct info
Dave Syer
@dsyer
May 10 2017 16:13
I guess it's just confusing then
Raise an issue if you want
Nastya Smirnova
@Aloren
May 10 2017 16:14
Json looks following way: "description": "Eureka discovery client is reporting failures to connect to a Eureka server",
"status": "UP",
"renewalPeriod": 30,
"failCount": 701414,
Okay, thanks. Will create an issue after debugging, cause in logs everything is good. :(
Felipe Adorno
@FelipeAdorno
May 10 2017 22:00
Hy all,
is possible use @Cachable in a Feing interface?
Mark Corkery
@KramKroc
May 10 2017 22:06
Hi folks. I’m doing some investigation into docker 1.13 & swarm with our existing Spring Boot/Spring Cloud system. We use Spring Boot Microservices with Eureka for service discovery of the services. I’m having problems with getting our services to see each other. Some of the services start on the ingress network, and others on the network I created for the stack. It’s pretty confusing and I’m not entirely sure if it’s an issue that I’ve created in docker stack setup or how I define my eureka instance id in Eureka. I’ve created a much smaller microservice system that reproduces the steps I followed and I wondered if someone could take a look to see if I’m missing something from the application/bootstrap yml definitions for the eurkea clients: https://github.com/KramKroc/nonzeroexit
Jonatan Ivanov
@jonatan-ivanov
May 10 2017 23:47