The dirty extents were not ready correctly for the block_status_64 return and as a result image server always reported the whole image range as dirty.
This commit fixes by normalizing the value returned by both block_status and block_status_64. It also adds warning whenever there is an issue getting the dirty extents and the code falls back to returning the full range as dirty.
Co-authored-by: Abhisar Sinha <63767682+abh1sar@users.noreply.github.com>
* KVM: add configurable MAC/IP script hook for static ARP/NDP and routes
Introduces a new agent.properties option `vm.network.macip.static`
(false by default) that makes BridgeVifDriver invoke on modifymacip.sh
on every NIC plug (VM start) and unplug (VM stop).
This is very useful in EVPN+VXLAN environments as it can reduce BUM
traffic. By setting static ARP/NDP entries bridges can be configured
using 'neigh_suppress on' as the ARP/NDP entries are already set
statically by CloudStack.
Setting 'neigh_suppress on' requires a manual change in the modifyvxlan.sh
script as this is not the default behavior.
* vxlan: In EVPN mode, disable ARP/NDP learning
FRR populates the FDB via BGP EVPN, so kernel data-plane learning is
redundant and counterproductive.
Static ARP (IPv4) and NDP (IPv6) entries are added on startup of the
Instance and remove on shutdown.
FRR populates the FDB/neighbor table via control plane, and neigh_suppress
tells the kernel bridge to use that information instead of flooding.
This will vastly reduce BUM traffic with static ARP/NDP entries.
Implements incremental backup support for the NAS backup provider on KVM, using QEMU dirty bitmaps and libvirt's backup-begin API. RFC: apache/cloudstack#12899.
For large VMs this reduces daily backup storage 80–95% and shortens backup windows from hours to minutes (e.g. a 500 GB VM with moderate writes goes from ~500 GB/day to ~5–15 GB/day after the initial full backup).
Signed-off-by: James Peru <jmsperu@gmail.com>
Co-authored-by: jmsperu <jmsperu@users.noreply.github.com>
Co-authored-by: Abhisar Sinha <63767682+abh1sar@users.noreply.github.com>
* NAS backup: resume paused VM on backup failure and fix missing exit
When a NAS backup job fails (e.g. due to backup storage being full or
I/O errors), the VM may remain indefinitely paused because:
1. The cleanup() function never checks or resumes the VM's paused state
that was set by virsh backup-begin during the push backup operation.
2. The 'Failed' case in the backup job monitoring loop calls cleanup()
but lacks an 'exit' statement, causing an infinite loop where the
script repeatedly detects the failed job and calls cleanup().
3. Similarly, backup_stopped_vm() calls cleanup() on qemu-img convert
failure but does not exit, allowing the loop to continue with
subsequent disks despite the failure.
This fix:
- Adds VM state detection and resume to cleanup(), ensuring the VM is
always resumed if found in a paused state during error handling
- Adds missing 'exit 1' after cleanup() in the Failed backup job case
to prevent the infinite monitoring loop
- Adds missing 'exit 1' after cleanup() in backup_stopped_vm() on
qemu-img convert failure
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* ci: retrigger workflow (flaky/stale shards)
* nasbackup.sh: keep cleanup() best-effort if domstate fails
With set -eo pipefail, a non-zero virsh domstate (libvirt unavailable or the
domain gone) in the vm_state assignment would abort cleanup() before the
rm/umount/rmdir. Append '|| true' so cleanup always runs to completion.
Addresses the review suggestion (Copilot, endorsed by @abh1sar and @weizhouapache).
Signed-off-by: James Peru <jmsperu@gmail.com>
---------
Signed-off-by: James Peru <jmsperu@gmail.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: jmsperu <jmsperu@users.noreply.github.com>
This PR introduces the initial implementation of Veeam integration support for KVM in CloudStack by adding a UHAPI-compatible server and image server components.
Veeam Backup & Replication interacts with virtualization platforms using its Universal Hypervisor API (UHAPI). To enable backup and restore workflows for CloudStack-managed KVM environments, this change introduces a UHAPI server that exposes CloudStack resources through a UHAPI-compatible interface.
In addition to the control plane APIs, an image server component is introduced to handle the data transfer operations required during backup and restore workflows.
The integration consists of two main components:
1. UHAPI Server (Control Plane) named CloudStack Veeam Control Service
A lightweight UHAPI server runs inside the CloudStack management server and exposes endpoints under:
/ovirt-engine
- /api - For APIs
- /sso - For authentication
- /services/pki-resource - For certificates
This server provides inventory discovery APIs required by Veeam and translates CloudStack resources into the structures expected by UHAPI.
The server:
- exposes infrastructure inventory
- handles authentication and session tokens
- maps CloudStack resources to UHAPI-compatible representations
2. Image Server (Data Plane) named CloudStack Image Service
A separate image server component is introduced to handle backup and restore data transfer operations.
This component:
- serves disk image data during backup
- receives image data during restore operations
- exposes endpoints used by Veeam worker components
- integrates with CloudStack storage to read and write VM disk data
The separation between both these components server ensures that:
- metadata APIs and control operations remain lightweight
- bulk image transfer operations are handled independently
Signed-off-by: Abhishek Kumar <abhishek.mrt22@gmail.com>
Co-authored-by: Abhisar Sinha <63767682+abh1sar@users.noreply.github.com>
Co-authored-by: abh1sar <abhisar.sinha@gmail.com>
Co-authored-by: Wei Zhou <weizhou@apache.org>
* Host HA code improvements
* Fix to not cancel VM HA items when Host HA is enabled & inspection in progress, and some code improvements
- When Host HA inspection in progress, the investigor returns the Host Status as Up which cancels the VM HA items
- Don't cancel the VM HA items, instead reschedule them to try again later
* Changes to consider Recovered/Available Host HA state along with the agent connection status to determine the Host HA inspection in progress or not, and some code improvements
* Fix domain parsing for GPU
* Add Display controller to GPU class check
this adds support for the amd instinct mi2xx accelorator crards in the discovery script.
Co-authored-by: Piet Braat <piet@phiea.nl>
* extension/proxmox: improve host vm power reporting
Add `statuses` action in extensions to report VM power states
This PR introduces support for retrieving the power state of all VMs on a host directly from an extension using the new `statuses` action.
When available, this provides a single aggregated response, reducing the need for multiple calls.
If the extension does not implement `statuses`, the server will gracefully fall back to querying individual VMs using the existing `status` action.
This helps with updating the host in CloudStack after out-of-band migrations for the VM.
Signed-off-by: Abhishek Kumar <abhishek.mrt22@gmail.com>
* address review
Signed-off-by: Abhishek Kumar <abhishek.mrt22@gmail.com>
---------
Signed-off-by: Abhishek Kumar <abhishek.mrt22@gmail.com>
* XenServer 8.4/XCP-ng 8.3: Support vTPM
* fix issue
* add log for windows 11 or other such guests OSs that require vtpm
* remove secure bootmode requirement
* Fix uefi setting on host for xenserver 8.4
This PR aligns the use of terminology, renaming VM / virtual machine references to 'Instance' and also capitalising the terms Templates, Network, Snapshot, User, Account in CloudStack APIs, error and log messages, events, tooltips, etc. Many typos, grammar and spelling mistakes were fixed, also terms like IPv4, VPN, VPC, etc. were properly capitalised. Some error messages were cleaned for better readability. The test cases, expecting some exception strings were adjusted accordingly.
Here is the wiki page, describing the changes in details:
https://cwiki.apache.org/confluence/display/CLOUDSTACK/Object+Naming+and+Title+Case+Convention
---------
Co-authored-by: Manoj Kumar <manojkr.itbhu@gmail.com>
Co-authored-by: Harikrishna <harikrishna.patnala@gmail.com>
This PR introduces console access support for instances deployed using Orchestrator Extensions, available via either VNC or a direct URL.
- CloudStack queries the extension using the getconsole action.
- For VNC-based access, the extension must return host/port/ticket details. CloudStack then forwards these to the Console Proxy VM (CPVM) in the instance’s zone. It is assumed that the CPVM can reach the specified host and port.
- For direct URL access, the extension returns a console URL with the protocol set to `direct`. The URL is then provided directly to the user.
- The built-in Proxmox Orchestrator Extension now supports console access via VNC. The extension calls the Proxmox API to fetch console details and returns them in the required format.
Also, adds changes to send caller details to the extension payload.
```
# cat /var/lib/cloudstack/management/extensions/Proxmox/02b650f6-bb98-49cb-8cac-82b7a78f43a2.json | jq
{
"caller": {
"roleid": "6b86674b-7e61-11f0-ba77-1e00c8000158",
"rolename": "Root Admin",
"name": "admin",
"roletype": "Admin",
"id": "93567ed9-7e61-11f0-ba77-1e00c8000158",
"type": "ADMIN"
},
"virtualmachineid": "126f4562-1f0f-4313-875e-6150cabeb72f",
...
```
Documentation PR: https://github.com/apache/cloudstack-documentation/pull/560
---------
Signed-off-by: Abhishek Kumar <abhishek.mrt22@gmail.com>
* scripts: fix external provision to use correct power state
The valid states are poweron and poweroff.
Signed-off-by: Rohit Yadav <rohit.yadav@shapeblue.com>
* strip string while processing powerstate for HyperV
Signed-off-by: Rohit Yadav <rohit.yadav@shapeblue.com>
* ignore warning that spills over to exten output string
Signed-off-by: Rohit Yadav <rohit.yadav@shapeblue.com>
---------
Signed-off-by: Rohit Yadav <rohit.yadav@shapeblue.com>
This feature adds the ability to create a new instance from a VM backup for dummy, NAS and Veeam backup providers. It works even if the original instance used to create the backup was expunged or unmanaged. There are two parts to this functionality:
Saving all configuration details that the VM had at the time of taking the backup. And using them to create an instance from backup.
Enabling a user to expunge/unmanage an instance that has backups.
This PR allows attaching of GPU devices via PCI, mdev or VF to an Instance for KVM.
It allows the operator to discover the GPU devices on the KVM host and create a Compute Offering with GPU support based on the available GPU devices on the host. Once the operator has created the Compute offering, it can be used by users to launch Instances with GPU devices.
The Extensions Framework in Apache CloudStack is designed to provide a flexible and standardised mechanism for integrating external systems and custom workflows into CloudStack’s orchestration process. By defining structured hook points during key operations—such as virtual machine deployment, resource preparation, and lifecycle events—the framework allows administrators and developers to extend CloudStack’s behaviour without modifying its core codebase.
* NAS B&R Plugin enhancements
* Prevent printing mount opts which may include password by removing from response
* revert marvin change
* add sanity checks to validate minimum qemu and libvirt versions
* check is user running script is part of libvirt group
* revert changes of retore expunged VM
* add code coverage ignore file
* remove check
* issue with listing schedules and add defensive checks
* redirect logs to agent log file
* add some more debugging
* remove test file
* prevent deletion of cks cluster when vms associated to a backup offering
* delete all snapshot policies when bkp offering is disassociated from a VM
* Fix `updateTemplatePermission` when the UI is set to a language other than English (#9766)
* Fix updateTemplatePermission UI in non-english language
* Improve fix
---------
* Add nobrl in the mountopts for cifs file system
* Fix restoration of VM / volumes with cifs
* add cifs utils for el8
* add cifs-utils for ubuntu cloudstack-agent
* syntax error
* remove required constraint on both vmid and id params for the delete bkp schedule command
This script can be used to use BGP+EVPN for VXLAN instead of multicast.
In order to use the script operators need to manually create a symlink:
cd /usr/share
ln -s modifyvxlan.sh cloudstack-common/scripts/vm/network/vnet/modifyvxlan-evpn.sh
Configuring BGP and EVPN is outside of the scope of CloudStack and is
up to the operator to configure.
This commit just adds this script to the main repository and does not
change any existing functionality.
This is a simple NAS backup plugin for KVM which may be later expanded for other hypervisors. This backup plugin aims to use shared NAS storage on KVM hosts such as NFS (or CephFS and others in future), which is used to backup fully cloned VMs for backup & restore operations. This may NOT be as efficient and performant as some of the other B&R providers, but maybe useful for some KVM environments who are okay to only have full-instance backups and limited functionality.
Design & Implementation follows the `networker` B&R plugin, which is simply:
- Implement B&R plugin interfaces
- Use cmd-answer pattern to execute backup and restore operations on KVM host when VM is running (or needs to be restored) - instead of a B&R API client, relies on answers from KVM agent which executes the operations
- Backups are full VM domain snapshots, copied to a VM-specific folders on a NAS target (NFS) along with a domain XML
- Backup uses libvirt feature: https://libvirt.org/kbase/live_full_disk_backup.html orchestrated via virsh/bash script (nasbackup.sh) as the libvirt-java lacks the bindings
- Supported instance volume storage for restore operations: NFS & local storage
Refer the doc PR for feature limitations and usage details:
https://github.com/apache/cloudstack-documentation/pull/429
Signed-off-by: Rohit Yadav <rohit.yadav@shapeblue.com>
Co-authored-by: Pearl Dsilva <pearl1594@gmail.com>
Co-authored-by: Abhishek Kumar <abhishek.mrt22@gmail.com>
Co-authored-by: Suresh Kumar Anaparti <sureshkumar.anaparti@gmail.com>
Extending the current functionality of KVM Host HA for the StorPool storage plugin and the option for easy integration for the rest of the storage plugins to support Host HA
This extension works like the current NFS storage implementation. It allows it to be used simultaneously with NFS and StorPool storage or only with StorPool primary storage.
If it is used with different primary storages like NFS and StorPool, and one of the health checks fails for storage, there is an option to report the failure to the management with the global config kvm.ha.fence.on.storage.heartbeat.failure. By default this option is disabled when enabled the Host HA service will continue with the checks on the host and eventually will fence the host
