[midPoint] Clustered Configuration - Communication error

[Pavol Mederly]

Matt,

most probably your keystores (keystore.jceks) on individual nodes are 
not synchronized. They must be the same in order for the authentication 
- or any password-related operations - to work reliably.

Take the keystore from the node where authentication works, and put it 
onto the other nodes.

Best regards,

-- 
Pavol Mederly
Software developer
evolveum.com

On 02/02/2022 00:53, Matt Widhalm via midPoint wrote:
> I am working on this one again. While trying to login to node_b I 
> am seeing the following error: Invalid username and/or password. This 
> is while using the administrator account. I have verified the password 
> is correct as I am able to login to node_a with the credentials.
>
> I have also tried to login using the emergency admin url and I am 
> seeing this error: Currently we are unable to process your request. 
> Kindly try again later.
>
> Has anyone run into similar issues while trying to set up a 
> clustered configuration?
>
> Any help would be greatly appreciated.
>
> Thank you,
> Matt Widhalm
>
>
> On Wed, Jan 12, 2022 at 11:47 AM Matt Widhalm 
> <matthewwidhalm at weber.edu> wrote:
>
>     Good morning. I am attempting to set up a clustered environment
>     utilizing 2 nodes. The issue I am running into is that while
>     node_a is in a running state, node_b is showing Communication
>     error. I have verified communication between the two docker
>     containers is working (I can ping node_b from node_a and vice
>     versa). Below is a section of my config.xml for node_a. node_b is
>     the same with the nodeId reflecting that it is node_b.
>
>             <repository>
>                 <type>native</type>
>     <jdbcUrl>jdbc:postgresql://<censored>:5432/midpoint_dev</jdbcUrl>
>     <jdbcUsername><censored></jdbcUsername>
>     <jdbcPassword><censored></jdbcPassword>
>     <missingSchemaAction>create</missingSchemaAction>
>     <baseDir>${midpoint.home}</baseDir>
>                 <asServer>true</asServer>
>             </repository>
>             <nodeId>node_a</nodeId>
>             <taskManager>
>                 <clustered>true</clustered>
>             </taskManager>
>
>     On the All Node page they are showing as:
>
>     Name: node_a
>     Status: Running
>     Contact: http://<CONTAINER ID>:8080/midpoint
>     Clustered: checked
>     Status message:
>
>     Name: node_b
>     Status: Communication error
>     Contact: http://<CONTAINER ID>:8080/midpoint
>     Clustered: checked
>     Status message: Node not known at this moment
>
>     I have tried to change the httpPort and url in the config.xml with
>     no change to the Communication error. I also have the containers
>     on their own network.
>
>     Any help would be appreciated.
>
>     Thank you,
>
>     Matt Widhalm
>
>     System Engineer
>
>     Weber State University
>
>
> _______________________________________________
> midPoint mailing list
> midPoint at lists.evolveum.com
> https://lists.evolveum.com/mailman/listinfo/midpoint
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.evolveum.com/pipermail/midpoint/attachments/20220202/f7549ccc/attachment.htm>