Docs feedback: Gitlab CE AWS setup
I have multiple questions what about the setup described at https://docs.gitlab.com/ce/install/aws/ what I don't understand. Can you please help me out?
- Configuration on "App Servers" need to be done one-by-one, right? There is no way to copy configuration during Auto Scaling Group operations? Which means a new server populated by the ASG will be unusable by default.
- An EBS volume (for storing Git data) need to be created for every node separately, and not as a shared (e.g. NFS storage)?
- Is it necessary to host the Gitaly component on different EC2 VM(s)? What's the difference / performance drawback if it runs on the same "App Servers"?
- For the Gitaly Server configuration you need to set storage paths like
/mnt/gitlab/default/repositories
. Is it the same path that needed to be set up here ? - The EBS volume need to be connected to the "App Server" EC2 instance, and not the "Gitaly Server" EC2 instance? Meaning on the "Gitaly Server" instance the storage path doesn't need to exist and contain the actual Git data?
- There is nothing to be found about how to set up "Read-only Gitaly Server" for the secondary node, and no options in the Gitaly documentation mentions anything about read-only mode.