An installation used to be this one: the deploy scripts and the Ansible defaults carried one site's domain, hypervisor and login, the Proxmox side and the database were set up by hand, and deploying meant copying scripts and a tree of roles beside the binary and keeping them in step with it. The binary now carries the roles, the Terraform module, the schema and the catalogue (embed.go), and installs itself: - hpcod setup: a full-screen installer for administrators. It asks where clusters are built, signs in to Proxmox before asking anything else (root's password once, or tokens), offers nodes and datastores from what Proxmox has, checks every field as it is typed, runs every step with its output live, checks the result and offers a test cluster. -yes runs the same steps from a site file; make deploy uses that. - hpcod verify: each fact checked by using it. - hpcod selftest: the smallest real cluster, by the portal's own path, a job run on it, then taken down. - Closed networks: hpcod serves HTTPS itself from a CA name-constrained to the site's domains, issuing and renewing its own certificate, and answers DNS for the portal and every cluster name. - A simulated Proxmox: Proxmox VE in QEMU on the control plane host, on a private bridge with NAT, for trying it without one. Every step reads back what it changed and is safe to run again. The site's values live in one file; group_vars/all/site.yml is generated from it, and the playbooks refuse to run without it. ci_user and -ssh-user no longer default to a personal account. scripts/B10, B20 and C20 are gone: their work is the setup steps. Proven end to end on a simulated Proxmox (Proxmox VE 9 in a VM on a real host): the install, the Proxmox step by the password path, the head template, verify READY, and a test cluster that ran srun on its compute node and was destroyed. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
55 lines
1.5 KiB
Go
55 lines
1.5 KiB
Go
package hpcod
|
|
|
|
import (
|
|
"io/fs"
|
|
"strings"
|
|
"testing"
|
|
)
|
|
|
|
// The tree must carry what a cluster is built from, and nothing that is
|
|
// one installation's own. Both directions are checked, because the failure
|
|
// in either is silent: a missing role surfaces on a head node at first
|
|
// boot, and a leaked inventory surfaces never.
|
|
func TestTreeCarriesWhatIsNeededAndNothingLocal(t *testing.T) {
|
|
var files []string
|
|
if err := fs.WalkDir(Tree, ".", func(p string, d fs.DirEntry, err error) error {
|
|
if err == nil && !d.IsDir() {
|
|
files = append(files, p)
|
|
}
|
|
return err
|
|
}); err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
have := map[string]bool{}
|
|
for _, f := range files {
|
|
have[f] = true
|
|
}
|
|
|
|
for _, want := range []string{
|
|
"ansible/site.yml", "ansible/golden.yml", "ansible/ansible.cfg",
|
|
"ansible/group_vars/all/main.yml",
|
|
"ansible/roles/selfconfig/files/hpcod-firstboot",
|
|
"terraform/modules/tenant/main.tf",
|
|
"terraform/modules/tenant/vendor-data.yaml.tftpl",
|
|
"terraform/envs/lab/main.tf", "terraform/envs/lab/variables.tf",
|
|
"db/001-schema.sql", "db/003-grants.sql",
|
|
"config/apps.yaml",
|
|
} {
|
|
if !have[want] {
|
|
t.Errorf("%s is not embedded", want)
|
|
}
|
|
}
|
|
|
|
for _, f := range files {
|
|
switch {
|
|
case f == "ansible/inventory/tenant.yml",
|
|
f == "config/sites.yaml",
|
|
f == "ansible/group_vars/all/site.yml",
|
|
strings.Contains(f, ".tfstate"),
|
|
strings.HasSuffix(f, ".tfvars"),
|
|
strings.Contains(f, "/.terraform"):
|
|
t.Errorf("%s is one installation's own file and must not be embedded", f)
|
|
}
|
|
}
|
|
}
|