failed to parse addresses from auto-join metadata in 1.19.1

[heatherezell]

@heatherezell hi, still encounted such error in 1.19.1 during fresh installation with the same configuration as 1.18.5, which was working fine in 1.18.5 failed to parse addresses from auto-join metadata: discover: label_selector: - equals in key's value, enclosing double-quote needed label_selector="value-with-=-symbol""

retry_join { auto_join = "provider=k8s label_selector=\"app.kubernetes.io/name=vault,component=server\" namespace=\"vault-cluster\"" auto_join_scheme = "http" }

Edit: After adding a space between comma and component in label_selector, the error is gone. Strangely, it is working fine in 1.18.5 and earlier version without the space.

Originally posted by @age9990 in #29863

[divyaac]

This should be fixed in the most recent version of Vault.

[peikk0]

I'm encountering the same issue with v1.19.1 and 1.19.2, but oddly only in 1 out of the 5 pods in the cluster. Adding spaces after each comma fixed it too, but that's still a bug that needs to be fixed.

[peikk0]

@divyaac this is a slightly different error from the one that was fixed in v1.19.1 (which I also ran into when upgrading to v1.19.0 previously), maybe the fix was incomplete.

[BigMacIT]

I seem to be having this issues running 1.19.1 also. I also can't find the fix in the release notes for 1.19.1, but it is in the changelog. Was trying to find out if something else may have changed, but I can't find for certain that it was fixed in a subsequent release. Can anyone point to the release that this is for sure fixed in?

[heatherezell]

Hello! (Hit enter too soon.) We have been able to reproduce this in our testing, and will be issuing a fix in the go-discover module. In the meantime, use a space as indicated - our test cases had a space in them, so this was missed. I cannot guarantee a timeframe for when the fix will be released, so watch this space. Thank you, everyone, for your patience. :)

[BigMacIT]

working with engineering, we found that it was triggered by an "initialization" so a node rebooting didn't have hte issue, but if you deleted the DB, or expanded the cluster, then it would surface. In our case this was the line in the stanza, so not sure the space has a play here.
auto_join = "provider=k8s namespace=vault label_selector="app=vault""

[heatherezell]

working with engineering, we found that it was triggered by an "initialization" so a node rebooting didn't have hte issue, but if you deleted the DB, or expanded the cluster, then it would surface. In our case this was the line in the stanza, so not sure the space has a play here. auto_join = "provider=k8s namespace=vault label_selector="app=vault""

You might be more specifically hitting the linked issue, #29863. They have the same root cause, I believe, so keep an eye on that issue for when it gets resolved. :)