After following, well verified steps as noted in > https://community.splunk.com/t5/Deployment-Architecture/How-to-move-the-SHC-deployer-to-another-host...
I was not able to successfully connect and test a push from the new deployer to the shcluster members. I received an error >>> Error while deploying apps to first member, aborting apps deployment to all members: Error while fetching apps baseline on target=https://host:8089: Non-200/201 status_code=401; {"messages":[{"type":"ERROR","text":"Unauthorized"}]}
Here are my steps:
1. copied the contents of /opt/splunk/etc/shcluster from the old deployer to the new deployer /opt/splunk/etc/shcluster
2) configured the new deployer [shclustering] stanza with the info from the old deployer [shclustering] stanza in /opt/splunk/etc/system/local server.conf
3) Updated conf_deploy_fetch_url in server.conf on each of the shc members
4) restarted the new deployer and a rolling restart on the shc members
5) did a test apply bundle and then received an error unauthorized.
I believe the issue could be with the pass4SymmKey (on the new deployer) not being the same as the pass4SymmKey on the SHC members.
I did a ./splunk show-decrypt --value <key> from the old deployer
[shclustering]
pass4SymmKey = <key>
shcluster_label = Company_shcluster1
I used the decrypted key as the key for the new deployer pass3SymmKey but ultimately I am not able to run a successful push.
Is there a way to recover these keys? The previous admin did not save the original secrets used to setup the deployer.
Any advice greatly appreciated.
Thank you
I actually was able to find the original secret that was used, so I got lucky.
After configuring the conf with the secret and a restart, test push was successful.
But thank you for the pointers, much appreciated!
When you are running
splunk show shcluster-status --verbose
splunk show kvstore-status --verbose
On those SHC nodes will you get a successful status with healthy SHC and KVStore?
Which version you have and which OS?
Easy to say after your issue, that I always prefer to use FQDN (usually CNAMEs) for all nodes to avoid this kind of issues when switch to another deployer, CM, LM or other....
r. Ismo
I actually was able to find the original secret that was used, so I got lucky.
After configuring the conf with the secret and a restart, test push was successful.
But thank you for the pointers, much appreciated!
Thank you for the additional info.
That is the weird part, I was running the show-decrypted on the original and I was getting garbage.
I just tried the show-decrypted again on the shclustering pass4SymmKey and I get just an "=" so I am not sure what is happening... but the password works...