Open epignosisx opened 4 years ago
I'd say it's all up to your configuration. Can you share?
Thanks for the quick reply. We are running it inside a container, here are the environment variables we are setting:
SPRING_CLOUD_CONFIG_SERVER_GIT_USERNAME=some-username SPRING_CLOUD_CONFIG_SERVER_GIT_PASSWORD=some-password SPRING_CLOUD_CONFIG_SERVER_GIT_URI=https://some-org@dev.azure.com/some-org/some-project/_git/app-config SPRING_CLOUD_CONFIG_SERVER_GIT_FORCE-PULL=true SPRING_CLOUD_CONFIG_SERVER_GIT_BASEDIR=/opt/app-config ENCRYPT_KEY=some-key MANAGEMENT_ENDPOINTS_WEB_EXPOSURE_INCLUDE=health
what version?
Seems odd that you have the username in the uri and env vars
what version?
2.1.2.RELEASE
Seems odd that you have the username in the uri and env vars
Indeed. I haven't noticed that. Let me check if there is a reason for it and remove it to see if it makes a difference.
That's an old version and I would suggest updating to Hoxton.SR8
Updated the git url to exclude the username, it did not make a difference. I'm in the process of updating the app to the latest version. I'll share the results.
We upgraded to the latest versions, but no luck: spring-boot-starter-parent from 2.1.5 to 2.3.4 spring-cloud-config-server from 2.1.2 to 2.2.5
It does not seem to be a problem with just the /actuator/health
endpoint, it also happens when requesting the configuration like /some-app/some-profile/some-label
@OlgaMaciaszek and @spencergibb any progress on this issue- I still see it happening 2020.0.4. any progress on that?
@spencergibb any update on that?
On springboot 3.3.1 SpringCloud 2023.0.2 Java21
cloud: config: server: bootstrap: ${CONFIGSERVER_BOOTSTRAP:true} encrypt: enabled: ${CONFIGSERVER_ENCRYPTENABLED:true} git: uri: https://gitlab.com/arctravel/arc-settlement/config/default-config.git basedir: ${CONFIGSERVER_BASEDIR}/default clone-on-start: ${CONFIGSERVER_CLONEONSTART} force-pull: ${CONFIGSERVER_FORCEPULL} username: ${GIT_USER} password: ${GIT_TOKEN}
@spencergibb is there anymore research I can provide to help in getting a fix for this one? This does limit our able to monitor for true production issues
Describe the bug We are running Spring Cloud Config with the Git backend and we've identified that when we call
/actuator/health
it's first making a call to the git repo and receiving a 401 (Unauthorized), and right away a call that succeeds. It appears as if the first request is not sending the credentials, but the second is.This doesn't seem right, am I missing something? I'm using a version a bit old (2.1.2.RELEASE), I looked through the issues but couldn't find any issue related to this.
Here is a screenshot of the outgoing calls that
/actuator/health
makes depicting the issue:Notice how this is not an isolated incident, but rather, each failed attempt is followed by a successful one: