[I] Configdrive/storage repository issue on XCP-ng [cloudstack]

["ingox (via GitHub)" <gi...@apache.org>]

ingox opened a new issue, #11984:
URL: https://github.com/apache/cloudstack/issues/11984

   ### problem
   
   Issue No1:
   In case of failed deployments there are unused storage repositories left behind.
   
   <img width="205" height="386" alt="Image" src="https://github.com/user-attachments/assets/bd04c8a7-ff78-494d-992b-3b078d543940" />
   
   `example:
   **xe vdi-list uuid=f16e0133-2074-49af-a393-1bb9fa007fb5**
   uuid ( RO)                : f16e0133-2074-49af-a393-1bb9fa007fb5
             name-label ( RW): i-2-24-VM.iso
       name-description ( RW):
                sr-uuid ( RO): **608c63b9-9c12-bf6a-8d28-f2bc7d01d343**
           virtual-size ( RO): 409600
               sharable ( RO): false
              read-only ( RO): true
   
   
   **xe sr-list name-label=i-2-24-VM-CONFIGDRIVE-ISO**
   uuid ( RO)                : 12b1e17a-fb1c-d99f-451c-baf035b822f2
             name-label ( RW): i-2-24-VM-CONFIGDRIVE-ISO
       name-description ( RW): 10.1.32.4:/acs/secondary/ref-trl-6095-x-Mu24/ref-trl-6095-x-Mu24-sec2/configdrive
                   host ( RO): <shared>
                   type ( RO): iso
           content-type ( RO): iso
   
   
   uuid ( RO)                : **608c63b9-9c12-bf6a-8d28-f2bc7d01d343**
             name-label ( RW): i-2-24-VM-CONFIGDRIVE-ISO
       name-description ( RW): 10.1.32.4:/acs/secondary/ref-trl-6095-x-Mu24/ref-trl-6095-x-Mu24-sec2/configdrive
                   host ( RO): <shared>
                   type ( RO): iso
           content-type ( RO): iso
   
   
   uuid ( RO)                : 222f3918-75e1-99f9-f7f2-e63d714c130f
             name-label ( RW): i-2-24-VM-CONFIGDRIVE-ISO
       name-description ( RW): 10.1.32.4:/acs/secondary/ref-trl-6095-x-Mu24/ref-trl-6095-x-Mu24-sec2/configdrive
                   host ( RO): <shared>
                   type ( RO): iso
           content-type ( RO): iso`
   
   
   
   As you see above just one storage repository is being used for the ISO file.
   This has no functional impact but leaves something unused behind.
   
   Issue No 2:
   Within a storage repository which name is specified to one particular VM you can see all the iso files from all the other VMs since they are located in the same secondary storage. Why not have one general storage repository like "VM-CONFIGDRIVE-ISO" to avoid both issues?
   
   
   
   ### versions
   
   ACS: 4.20.1
   XCP-ng: 8.2
   
   ### The steps to reproduce the bug
   
   1. Build a couple of VMs where you have limited resources available on the hypervisors.
   2. 
   
   <img width="491" height="308" alt="Image" src="https://github.com/user-attachments/assets/4b7cf30d-6d40-45e4-ae5c-4befc9d81972" />
   
   3. The hypervisor still has been selected for deployments but failed from hypervisor side
   4. Several attempts have happened
   ```
   cat  /var/log/cloudstack/management/management-server.log | grep job-117 | grep  'VM start attempt'
   2025-10-28 14:22:19,361 DEBUG [c.c.v.ClusteredVirtualMachineManagerImpl] (Work-Job-Executor-42:[ctx-a1bf013f, job-117/job-118, ctx-520420a9]) (logid:67fb25c7) VM start attempt #1
   2025-10-28 14:22:24,301 DEBUG [c.c.v.ClusteredVirtualMachineManagerImpl] (Work-Job-Executor-42:[ctx-a1bf013f, job-117/job-118, ctx-520420a9]) (logid:67fb25c7) VM start attempt #2
   2025-10-28 14:22:25,371 DEBUG [c.c.v.ClusteredVirtualMachineManagerImpl] (Work-Job-Executor-43:[ctx-34843ad3, job-117/job-119, ctx-605dae51]) (logid:67fb25c7) VM start attempt #1
   2025-10-28 14:22:28,159 DEBUG [c.c.v.ClusteredVirtualMachineManagerImpl] (Work-Job-Executor-43:[ctx-34843ad3, job-117/job-119, ctx-605dae51]) (logid:67fb25c7) VM start attempt #2
   ```
   5. For each attempt a storage repository has been created. Every time an ISO file as well. On failed attempt the ISO file has been removed but the storage repository not. 
   6. Once the VM finally got deployed you'll end up with multiple storage repositories, used and unused once with the same name.
   
   
   ### What to do about it?
   
   Option 1: Besides removing the ISO file do also remove the storage repository on failed deployments.
   Option 2: Use one general storage repository for all ISO files.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscribe@cloudstack.apache.org.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org