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

12th
Aug 2016
bitsofinfo
@bitsofinfo
Aug 12 2016 00:54
This is a bit of a blocker for me right now: spring-cloud/spring-cloud-netflix#1252
soky srm
@sokie
Aug 12 2016 07:11
Hi guys, if I use sleuth with a rabbitmq stream. If the rabbit service goes down, will the logs be buffered and sent once the service is up, or are they lost"?
soky srm
@sokie
Aug 12 2016 07:42
ok, analysis with rabbitmq management shows it doesn't
any chance that's configurable?
Thibaud Lepretre
@kakawait
Aug 12 2016 13:23

@spencergibb do you think is technical possible extend <serviceId>.ribbon.* properties to add relaxed binding capability. Actually (example) ConnectTimeout is case sensitive. Moreover is not possible to override properties using external configuration as explained http://docs.spring.io/spring-boot/docs/current/reference/html/boot-features-external-config.html, I tried with -D<serviceId>.ribbon.ConnectTimeout=9000 it does not work (replace <serviceId> with existing service id).

I never look how relaxed binding work internally so maybe is not technically possible with such dynamic properties (dynamic part <serviceId>)

(and maybe other properties from other products like Hystrix)
pradeepkusingh
@pradeepkusingh
Aug 12 2016 15:57
@spencergibb Eureka is wrongly claiming that selfpreservation is false
EMERGENCY! EUREKA MAY BE INCORRECTLY CLAIMING INSTANCES ARE UP WHEN THEY'RE NOT. RENEWALS ARE LESSER THAN THRESHOLD AND HENCE THE INSTANCES ARE NOT BEING EXPIRED JUST TO BE SAFE.
i am seeing that we have proper setting
eureka.server.enableSelfPreservation: true
why any idea or I am using wrong proeprty ?
Spencer Gibb
@spencergibb
Aug 12 2016 15:58
that message is that self preservation is turned on and is currently active. It’s protecting itself.
pradeepkusingh
@pradeepkusingh
Aug 12 2016 15:59
hmm..strange sometimes this shows this message
RENEWALS ARE LESSER THAN THE THRESHOLD. THE SELF PRESERVATION MODE IS TURNED OFF.THIS MAY NOT PROTECT INSTANCE EXPIRY IN CASE OF NETWORK/OTHER PROBLEMS.
so I am confused...
just scratching my head.. with different messages
Spencer Gibb
@spencergibb
Aug 12 2016 16:00
that message means eureka.server.enableSelfPreservation: false and the threshold has been reached where it WOULD have protected itself.
check /env to make sure
pradeepkusingh
@pradeepkusingh
Aug 12 2016 16:00
i checked
that was the first thing
applicationConfig: [file:./application.yml]: {
server.port: "${PORT:8761}",
eureka.client.registerWithEureka: false,
eureka.client.fetchRegistry: false,
eureka.client.serviceUrl.defaultZone: "http://localhost:8761/eureka/",
eureka.instance.hostname: "${vcap.application.uris[0]:localhost}",
eureka.instance.leaseRenewalIntervalInSeconds: 3,
eureka.instance.leaseExpirationDurationInSeconds: 30,
eureka.instance.metadataMap.instanceId: "${vcap.application.instance_id:${spring.application.name}:${server.port:8080}}",
eureka.server.enableSelfPreservation: true
},
last line has that value as "true" and I made sure not multuple values in any other location of files
Spencer Gibb
@spencergibb
Aug 12 2016 16:02
don’t know then
pradeepkusingh
@pradeepkusingh
Aug 12 2016 16:08
hmm..thanks..
David Welch
@dwelch2344
Aug 12 2016 17:40
Question: if the service is failing in consul, on it's HTTP check, will ribbon still route requests to it until it's registered? Because that feels like what we are seeing
David Welch
@dwelch2344
Aug 12 2016 17:58
Ty :)
Spencer Gibb
@spencergibb
Aug 12 2016 18:15
@dwelch2344 lot’s of folks asking about that kind of thing for blue green deploys.
The question is often asked how netflix does it. The bring up new instances in 'out of service’ status. The change new instances from 'out of service’ to ‘up’ and old instances from ‘up’ to out of service. Old instances remain functioning to take requests until the caches eventually drop all old instances.