I'd like to create a local yum repository for my organization with semi-custom RPM spec files to handle the upgrade procedure for Splunk Universal Forwarders (and Splunk Enterprise to some extent). In limited testing this seems to work fine. From a deployment perspective it seems like it would be a lot easier to maintain than integrating custom installation scripts, Puppet, or some other deployment tool.
I'm curious if anyone else has implemented this and encountered any pitfalls that I'm overlooking. In my mind it should be just as straightforward as it sounds:
- All of our configurations and applications will be in directories untouched by the upgrade procedure, so there's no threat to overwriting anything of importance
- $SPLUNK_HOME/var isn't contained in the RPMs/tarballs, so there's no risk to data being overwritten
Is there anything that I'm overlooking that would make this a less desirable solution than using another deployment tool?
↧