Releases: lae/ansible-role-proxmox
v1.9.0
Somewhat hefty feature release that includes some longstanding work. Major changes are as follows:
- VFIO device stubbing and other configuration can now be managed. See the PCIe Passthrough section for details.
- Metric servers can now be defined.
- LDAP/AD domains and other realms can now be configured.
What's Changed
- pvesh: Fix invalid string escape sequences by @tbabej in #244
- fix: Replace deprecated ansible.netcommon.ipaddr by @btravouillon in #242
- Support for adding authentication realms to domains.cfg by @jfpanisset in #95
- Support for authentication realms configuration by @mrtwnklr in #220
- fix minimum host count in inventory group before enabling pve-cluster by @tbasset in #224
- VFIO/PCIe Passthrough configuration support by @ol3d in #247
- Fix minimum host check for clustering by @lae in #250
- Support for CIFS storages, namespacing in PBS, and ranking interfaces for corosync by @edv-pi in #251
- Metric Servers management in Proxmox by @thystips in #262
- Install Amplify Workflow by @amplify-security in #266
- Run Vagrant in CI by @lae in #269
- Allow version pinning of the Proxmox kernel by @lae in #267
- Release 1.9.0 by @lae in #263
New Contributors
- @tbabej made their first contribution in #244
- @mrtwnklr made their first contribution in #220
- @tbasset made their first contribution in #224
- @ol3d made their first contribution in #247
- @edv-pi made their first contribution in #251
- @thystips made their first contribution in #262
- @amplify-security made their first contribution in #266
Full Changelog: v1.8.1...v1.9.0
Contributors
Assets 2
v1.8.1
Contributors
Assets 2
v1.8.0
Release notes subject to change (using automatic release notes here) to be more useful, but this adds PVE 8 support and updates Ceph default repo.
There are probably upgrade steps you need to follow from https://pve.proxmox.com/wiki/Ceph_Pacific_to_Quincy, and possibly other steps as well depending on what PVE features you're using.
Not sure if this will be installable from the new Ansible Galaxy, so just in case, refer to lae/ansible-role-netbox#174 for some advice.
Managed to get this in before Sana finished her stream, huh. Also happy birthday, Hoshimiya ichigo.
What's Changed
- Idempotency fixes for proxmox_storage by @lae in #204
- Fix parsing of dpkg-query for old kernel packages by @lae in #206
- Update README.md by @pgagarinov in #209
- Update collect_kernel_info.py by @Rijo-ru in #216
- feature: add btrfs as valid storage type by @lexxxel in #229
- Add Proxmox Backup Server storage type by @lae in #169
- feature: add debian 12 bookworm and proxmox 8 compatibility by @lexxxel in #230
- updated default ceph version to quincy by @wu3rstle in #215
- Fix host used to configure HA groups by @btravouillon in #219
- Release 1.8.0 by @lae in #235
New Contributors
- @pgagarinov made their first contribution in #209
- @Rijo-ru made their first contribution in #216
- @lexxxel made their first contribution in #229
Full Changelog: v1.7.1...v1.8.0
Contributors
Assets 2
v1.7.1
Mostly bug fixes and some minor improvements/additions. Proxmox pools can now be configured.
Changes/Fixes
#116 Initial support for EFI-based systems. Should help with running the role on PVE installed via ISO somewhat, but this is still not supported.
#187 Use more correct SSH service name
#188 Fix check mode for deployments using Ceph
#189 Support for configuring Proxmox pools
#190 Ignore kernel module directories that have no installed kernel when doing kernel detection
#191 Fix ZFS modprobe options from not taking effect
#194 Fix duplication issue when using multiple CephFS mounts
#195 Fix Ceph deploy issue when there are no OSDs
#200 Deprecate legacy clustering variables as stated in release 1.6.9
Full Changelog: v1.7.0...v1.7.1
v1.7.0
Major feature release, adding support for PVE 7.x deployments. PVE 5.x can not be deployed with this version (6.x is fine).
Due to an oversight to how PVE merges known_hosts when adding nodes to a cluster, earlier versions of this role resulted in a convoluted /etc/pve/priv/known_hosts. This has been fixed, but you may wish to modify that file by hand to remove duplicates and comments. This is mainly a cosmetic issue and (as far as I know) doesn't affect PVE's operation.
Changes/Fixes
#114 ZFS pool and ZFS volume creation support
#125 Enable cache updating on PVE installation apt task
#155 PVE 7.x support and 5.x deprecation
#168 Remove unsupported implementation of Let's Encrypt certificate generation
#172 Subscription patches removed.
#173 Ensure hosts with PVE already installed can join clusters
#176 Fix idempotency verification of Ceph crush map modification tasks
#179 Fix module naming conflict on ceph_volume
#183 Allow rebuilds of first host in a cluster inventory
Full Changelog: v1.6.9...v1.7.0
v1.6.9
Final release for PVE 5.x. A couple bug fixes/improvements and Ceph improvements.
Please take this opportunity to upgrade your cluster to PVE 6.x (and then to 7.x if you can) if you have not done so already, as 5.x/Stretch is EOL.
#121 Add autoscale mode for Ceph pool configuration
#122 pve_ceph_mgr_group role variable to specify group of hosts to deploy the Ceph manager daemons on
#123 Correct list of required attrs for RBD storages
#124 Add the ability to disable hosts file management with pve_manage_hosts_enabled role variable
#128 Add CephFS storage type
#130 Ensure Ceph monitors only deploy on hosts in pve_ceph_mon_group
#131 Add pve_ceph_nodes role variable to limit which nodes Ceph is installed on.
#135 Replace fetch tasks with slurp and identify kernel in check mode
#136 Subscription patch fix (although this is outdated by the time of this release)
#139 Simplify patching of checked_command function
#140 (and subsequent fix in fbe315c) Manage /etc/hosts, even when clustering is disabled
#146 Throttle rebooting of cluster nodes when upgrading kernels (with optional pve_reboot_on_kernel_update_delay role variable)
v1.6.8
Mostly a bug fix release.
#112 Addition of a pve_roles role variable for configuring access roles in Proxmox
#115 Idempotency fix for ACL configuration in PVE 6
4d55dd5 Subscription patch fix for PVE 6.3 (#118)
2c0f021 Fix for deploying multiple standalone PVE nodes within the same inventory group (#117)
v1.6.7
Mostly improvements to management of Ceph clusters under PVE and some minor enhancements/fixes.
#100 Support modifications of the CRUSH map
#101 Allow CRUSH rule types to be specified
#103 Support creation of encrypted OSDs
#105 Ensure gpg is installed
#108 Skip modification of /etc/hosts and PVE::CLI::pvecm if clustering is not enabled
9e8e49e Update subscription patch for PVE 6.2-15
v1.6.5
v1.6.4
Hot fix release for PVE 6.x.
Changes/Fixes
91fc05f Patches the pvecm CLI library to fix cluster join failures (i.e. retain backwards compatibility). This change will be reverted once upstream has pushed an update with this patch. See #91 for more information.
#90 pvelocalhost is no longer inserted into system host files.
#89 pve_manage_ssh can now be used to toggle SSH configuration management. You can use this if you use a role to manage the entire SSH config file - but please note that cluster configuration may become a bit more difficult (as this role inserts config blocks based on your inventory).