All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- include diskless assignments in the GUI's "VM Disks"
- Performance regression (querying the LINSTOR controller too often) on status calls (requires LINSTOR >= 1.23.0)
- switch to cluster wide storage view instead of node-local view (requires LINSTOR >= 1.21.1)
- depend on a recent enough LINSTOR version (we depend on the satellite, but require the controller API)
- Properly handle multiple storage pools in one resource group. This allows to confine failure domains to a single back-end device as described here.
- Container clone from snapshot: only create "temporary" DRBD volume if necessary; remove this temporary DRBD resource when the snapshot gets removed.
redundancy
: The setting was unused since a long time (i.e., when we introduced LINSTOR resource groups). Remove some leftover dead code.controllervm
: Should not be used any longer, please usedrbd-reactor::promoter
to set up a HA LINSTOR Controller
- default for
statuscache
: This was set to0
(i.e., disabled) by default. Enable this cache by default and set it to 1 Minute. - default for
preferlocal
: Set default toyes
.
- Support for APIVER10
- Support for APIVER9
- Allow https communication to the API via
apicrt
,apikey
,apica
- UEFI disks on ZFS
- Set User-Agent to
linstor-proxmox/$version
- Add
statuscache
storage.cfg option
- Support for multiple LINSTOR controllers
- Support for APIVER7/8
- Support for APIVER6
- increase debian/changelog
- Remove temporary resource after
vzdump
in case of backup via a snapshot.
- Snapshot rollback to the last snapshot
- Backups via snapshots (i.e., snapshot support in
map_volume()
/activate_volume()
)