#ansible
What are currently the best #ansible resources to deploy #ceph these days in a fresh env?
There are a lot of outdated or non-collection/role like resources out there, some only have partial "bootstrap" scripts, other say "please install first and the manage via this one".
- cephadm-ansible
- stackhpc/ansible-collection-cephadm
- ceph/ceph.automation
etc.
It just got merged… You can now easily manage storage in #Proxmox with #Ansible.
Simply add CephFS, NFS, iSCSI or Proxmox Backup Server to your cluster with a simple Ansible task :)
#proxmox #proxmoxve #ansible #opensource #contributing #foss #automation #devops
Oh please!!!!
I wanted to add a smoke-test/integration test for my Ansible FreeBSD connection plugin against a test-server and the GitHub Actions don't support outgoing IPv6 in 2025?! WTF!
Tempted to move my remaining stuff off of GitHub and migrate my CI/CD to @Codeberg Woodpecker
I refuse to use IPv4 for that, just because of GitHub.
So, der Kollege und ich haben diese Woche reihenweise uralte Server abgeschaltet, andere Server gepatcht, Schutzmechanismen hochgezogen etc.
OpenVAS scannt über das Wochenende nach Sicherheitslücken und wir erwarten, dass das Ergebnis am Montag eeetwas weniger rot ist 
Automatisches Installieren von Updates, Konfiguration von SELinux, lokalen Firewalls etc. habe ich per Ansible auf dem Schirm.
Proxmox fully automated! From ClickOps to Code: Automated. Audited. Revisioned. Repeatable.
Starting from the base by automating:
- Cluster initialization
- Cluster join
- Storage Integration
- Proxmox Backup Server Integration
- SDN Networks (different ones for pros/dev)
- Guest Resources utilizing the cluster infrastructure
#Proxmox #PVE #Pbs #ProxmoxBackupServer #opensource #Automation #Ansible #python #devops #terraform #cicd #pipeline #cluster #nfs #iscsi
IT‑Spezialist:in für digitale Langzeitverfügbarkeit gesucht!
Das Zuse‑Institut Berlin (ZIB) sucht Unterstützung für das digitale Langzeitarchiv „EWIG“. Du arbeitest mit #OpenSource Tools wie Archivematica an Archivierungs‑Workflows, CI/CD, Deployment, Monitoring und Datenkuratierung.
Techstack: #Python, #Linux, Container, idealerweise #Ansible & CI/CD-Tools
Vollzeit · befristet (3 Jahre) · EG 12 TV‑L · Berlin
Bewerbung bis 31.08.2025
https://www.zib.de/jobadvertisement/1725-it-spezialist-fur-digitale-langzeitverfugbarkeit-wmd
I build a custom #Ansible facts module which works fine when I add something like this to ansible.cfg:
facts_modules = smart,name_of_my_custom_fact_module
However this module makes sense for a certain role only.
Is there a way to limit calling this module to a certain role?
For admins who speak YAML like a native: Passbolt’s Ansible playbook documentation is available.
This installation method is recommended for experienced users proficient with Ansible and requires a clean Debian 12 or Red Hat 9 server.
Full step-by-step guide here → https://www.passbolt.com/docs/hosting/install/pro/ansible/
GitHub repo → https://github.com/passbolt/passbolt-ansible-playbooks
I woke up with no other ideas on how to solve the ansible_managed thing than with a filter. (A lookup plugin would have also done it but I see no advantages in it.)
Proof of concept at the tail end of https://jpmens.net/2020/09/29/using-ansible-managed/
I mentioned recently that the configurability of {{ ansible_managed }} is being removed for #ansible 2.23. The reasoning appears to be “can be set in an inventory or other vars source”.
Anybody have any clever ideas on how to accomplish template filename modification time without prior invocation of a module or two, preferably also without having to create a custom lookup/filter plugin? (The template filename is in {{ template_path }}, {{ template_uid }} )
I’m out of ideas.
2/2
When we originally invented #ansible’s “ansible_managed” variable, we made it configurable. Before the default was broken a few years ago, it was roughly:
ansible_managed = Ansible managed: {file} modified on %Y-%m-%d %H:%M:%S by {uid} on {host}
One of the cool bits (I say this as its inventor) was that such an entry in an ansible.cfg caused the templated value to have the filename of the template source {file}, it’s owner {uid}, and it’s modification time in the strftime() patterns.
1/2
Actually on the topic of feature flags, I'm curious if they exist/what people are doing at the infra / #IaC / #GitOps level.
At one point I was thinking "can I hookup like launchdarkly into #Puppet's Hiera" to handle phased rollouts of things.
More recently, want the same thing for my #FluxCD stuff. Higher level than like Flagger I think, as in gradual rollout of Deployments over many clusters, than gradual rollout of Pod within a deployment.
Kinda like #Ansible's strategies and it's "max_fail_percentage" where it'll halt.
#Ansible scheduled a virtual #Meetup on October 2nd.
Sign up here:
https://www.meetup.com/ansible-virtual-meetups/events/310258600/