Hi all, We have a search head cluster that is spread across four locations - two sets of virtualised search heads that were the original cluster, and two sets of physical search heads that are the new cluster (the aim being that once the new search heads have settled in we will be removing the virtualised search heads). At present, load balancing for GUI access is only to the virtualised search heads. The issue we are having is that when one of the new physical search heads is made captain all of the dashboard panels in every app display "Action forbidden". The cluster is also unable to determine the cluster status when one of the physical search heads is captain, but can when one of the virtual search heads is captain. Everything else appears to be working correctly - searches appear to be being scheduled and distributed correctly, whether it is one of the physical or virtual search heads as captain. Clustering ports are open - 8089 for management, 8191 for kvstore, plus our replication port Authentication data is appropriately present on the new search heads, including the account I am testing with Has anyone come across a similar issue in the past or have ideas for a solution? Thanks, Mike ... View more

Hi, We run a distributed architecture in which we have organisations connecting in to our license master, and one organisation has been having trouble connecting. We checked the usual network troubleshooting and it seems fine, and I eventually found a line in the splunkd log suggesting an SSL error: "WARN HttpListener - Socket error from X.X.X.X:Y while idling: error:140760FC:SSL routines:SSL23_GET_CLIENT_HELLO:unknown protocol"   So I didn't think license management traffic actually had an SSL option in the first place, and can't see any reference to it in Splunk's SSL documentation. Has anyone seen any similar errors or have any idea how it might be resolved?   Thanks, Mike ... View more