Indexer registering with Manager, assuming managers IP address

[mike1002]

I am developing a splunk setup using docker image and Podman.  I am trying to setup 2 indexers along with an indexer manager.  Each container will run on separate rhel vm.  I successfully set up the Manager.  I then go to register the indexer as a peer and enter in the vm host IP of the manager and successfully register the indexer as a peer.  

When I reboot and check the indexer manager, it shows the indexer peer is up and up but shows the ip address of the manager container for the indexer peer?  When I try to add another indexer it does the same thing and will not let me add another indexer.  I have tried statically assigning IPs and confirmed all IPs are different etc.  I wasn't sure If anyone has ran into this issue.

All vm hosts are on the same subnet and can communicate.  Firewall off and selinux off. Using 9887 as rep port and 8089 as manager comms port. 

I am running as rootless outside and root inside.  It has to be a permission or file that I am missing.  I set it up as root:root and it works perfect.  Any ideas I appreciate it. 

[isoutamo]

Hi
Why you don’t use e.g. Splunk Operator for Kubernetes or Splunk’s docker version?
https://splunk.github.io/splunk-operator/ and https://github.com/splunk/docker-splunk
r. Ismo

[mike1002 · Accepted solution]

For unspecified reasons I cannot use K8 yet.  For this build I was attempting to utilize Splunk docker image with Podman in a rootless mode.  Didn't know if anyone had any experience with this.  It seems to be a privilege issue since root:root registers fine with vm host IP of the indexer peer.  For some reason rootless:root does not and assumes the container IP of the manager container.