How are people using the deployment server for deploying the Windows and Unix Technology-Addons?
The TA's are distributed as .tar.gz archives and the docs for serverclass.conf say that the appFile property supports directories, .tar and .tgz files.
I can get the deployment server working by extracting the directories/files from the archives and the clients pick these up fine.
Is this the way to do it or do people deploy the archives without extracting them and if so how?
It looks like this was my own fault. The machineTypesFilter I was using wasn't being applied because there wasn't a corresponding whitelist/blacklist entry.
I did note that when the UF unpacked the TA archive the Splunk_TA_nix appdirectory structure looked weird, it had another Splunk_TA_nix sub-directory underneath. I've since reverted to unpacking the TA archives and deploying them that way.
It looks like this was my own fault. The machineTypesFilter I was using wasn't being applied because there wasn't a corresponding whitelist/blacklist entry.
I did note that when the UF unpacked the TA archive the Splunk_TA_nix appdirectory structure looked weird, it had another Splunk_TA_nix sub-directory underneath. I've since reverted to unpacking the TA archives and deploying them that way.