InvalidTokenException due to Refresh Token not refreshed automatically

[dietzsch]

In the past we used CF Java client version 3.13.0.RELEASE in our Spring Boot application. We are using a configuration to create the CloudFoundryClient as bean, which we are using later on to call the REST API. The REST API is triggered via scheduler once a week. This setup has worked for numerous months.
When we updated the CF Java client to version 3.21.0.RELEASE we get the following error after some time:

2020-05-27 00:25:57 ERROR org.springframework.scheduling.support.TaskUtils$LoggingErrorHandler Unexpected error occurred in scheduled task org.cloudfoundry.reactor.util.Operator$ResponseReceiver$InvalidTokenException: null Suppressed: reactor.core.publisher.FluxOnAssembly$OnAssemblyException: Assembly trace from producer [reactor.core.publisher.MonoCacheTime] : reactor.core.publisher.Mono.checkpoint(Mono.java:1877) org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.token(AbstractUaaTokenProvider.java:255) Error has been observed at the following site(s): |_ Mono.checkpoint ⇢ at org.cloudfoundry.reactor.tokenprovider.AbstractUaaTokenProvider.token(AbstractUaaTokenProvider.java:255)

After some research this error seems to be related to the expiration of the Authorization/Refresh Token, where the TokenProvider is no longer refreshing the Refresh Token automatically. This seems to be a behaviour change between the two CF Java client versions. My understanding is that when using a TokenProvider (e.g. PasswordGrantTokenProvider) the credentials are provided once and the TokenProvider is taking care of refreshing the Refresh/Authorization Token, when necessary. Especially as a user of the client is not aware of when the toke is expiring.

[twoseat]

HI @dietzsch - it's something we're aware of, and are looking to release something to handle it very soon. It appears that the original behaviour was accidental, and continued working only because certain possible error conditions happened to arise rarely. So we're working on something that will work because we mean it to, and will be more resilient than the original.

[twoseat]

@dietzsch As you can see I've pushed a fix. It's caught up in some CI issues right now, but once it makes it through please grab the build snapshot and let me know if you have any problems.

[dietzsch]

@twoseat Thank you for the quick fix. Will try the latest snapshot build 3.25.0.BUILD-20200611.192423-4 that contains the fix. It will take some time on our end to verify the fix, but I will update the ticket as soon as we have results.

[jvalkeal]

Just tried it and works ok, thx!

[dietzsch]

@twoseat Meanwhile we were also able to verify the fix. Is there is already any plans when 3.25.0.RELEASE will be released?

[twoseat]

@dietzsch It's already released, I just forgot to do the release notes! Fixing that now...

[dietzsch]

@twoseat Wow. Did not expect to get the release that fast. :-)