This file identifies uniquely the repository/server and is used to indicate the ''primary node'' on a cluster (more than one ALM server properly installed to point to the SAME DB/Repos resources).
If you choose ''second node'' on install and point to the SAME existing DB/qcsiteadmin used by another ALM server AND do NOT point to the same repository (i.e. via a network share), it will FAIL because it cannot hash the REPID correctly against what is stored in the qcsiteadmin_db indicating the primary node's repid.
I had a situation recently where someone copied the repository from PROD to TEST and carried with it the repid file and it messed up the install claiming mismatch about repid and primary node - even though we were installing as a stand-alone.
for some reason, setting the ''repository validation'' flag to false on install did not help either.
Our solution was to rename the repository copy, reinstall and allow QC to create a new/blank repository. This created a NEW repid and it was happy.
When finished and running, we stopped the service and COPIED the various project folders into the proper positions under the repositoryqc folder. We first created the representative ''domain folders'' and copied the individual project folders in since they were so large instead of copying entire domain folders.