[FR] Add option to exclude disk sync for specific VMs

README.md

@@ -1,4 +1,3 @@
-
 # NetBox-Sync
 
 This is a tool to sync data from different sources to a NetBox instance.
@@ -295,6 +294,17 @@ Check out the documentations for the different sources
 * [vmware](https://github.com/bb-Ricardo/netbox-sync/blob/main/docs/source_vmware.md)
 * [check_redfish](https://github.com/bb-Ricardo/netbox-sync/blob/main/docs/source_check_redfish.md)
 
+### Filtering
+netbox-sync provides various filtering capabilities to control what objects are synced from sources to NetBox:
+
+1. **General VM filtering**: Use `vm_include_filter` and `vm_exclude_filter` to include or exclude VMs by name.
+2. **Tag-based VM filtering**: Use `vm_exclude_by_tag_filter` to exclude VMs with specific vCenter tags.
+3. **Partial information filtering**:
+   - Use `vm_exclude_disk_sync` to exclude disk synchronization for VMs matching specific name patterns.
+   - Use `vm_exclude_disk_sync_by_tag` to exclude disk synchronization for VMs with specific vCenter tags.
+
+These filters allow for fine-grained control over what information is synchronized, helping to avoid clutter in the change log from temporary or backup-related disk changes.
+
 If you have multiple vCenter instances or check_redfish folders just add another source with the same type
 in the **same** file.
 

docs/source_vmware.md

@@ -107,3 +107,25 @@ Primary IPv4/6 will be determined by interface that provides the default route f
 
 **Note:**<br>
 IP address information can only be extracted if guest tools are installed and running.
+
+### Filtering VM Disk Information
+VM disks are synchronized between vCenter and NetBox. Since NetBox 3.7.0, virtual disks are tracked as separate objects 
+linked to VMs. In some scenarios, such as when temporary disks are attached to VMs during backup operations 
+(e.g., "Independent-nonpersistent" disks from Veeam), you might want to exclude these changes from synchronization 
+to avoid cluttering your NetBox change log.
+
+You can use the following filter options to exclude disk synchronization for specific VMs:
+
+1. **`vm_exclude_disk_sync`**: A regex pattern matching VM names where disk synchronization should be excluded.
+   ```ini
+   vm_exclude_disk_sync = backup-.*, veeam-.*
+   ```
+
+2. **`vm_exclude_disk_sync_by_tag`**: A comma-separated list of vCenter tags. VMs with any of these tags will have 
+   their disk information excluded from synchronization.
+   ```ini
+   vm_exclude_disk_sync_by_tag = backup-vm, veeam-job
+   ```
+
+When a VM matches these filters, it will still be synchronized to NetBox with all its other information 
+(CPU, memory, interfaces, IP addresses, etc.), but changes to disk information will be ignored.

module/sources/vmware/config.py

@@ -116,7 +116,22 @@ def __init__(self):
                          """,
                          config_example="tag-a, tag-b"
                          ),
-
+            ConfigOption("vm_exclude_disk_sync",
+                         str,
+                         description="""defines a comma separated list of VM names (regex) where disk synchronization 
+                         will be excluded. A VM matching this filter will still be synced to NetBox,
+                         but its disk information won't be updated.
+                         """,
+                         config_example="backup-.*, temp-.*"
+                         ),
+            ConfigOption("vm_exclude_disk_sync_by_tag",
+                         str,
+                         description="""defines a comma separated list of vCenter tags which (if assigned to a VM)
+                         will exclude this VM from disk synchronization. A VM with this tag will still be synced 
+                         to NetBox, but its disk information won't be updated.
+                         """,
+                         config_example="backup-vm, veeam-job"
+                         ),
             ConfigOptionGroup(title="relations",
                               options=[
                                 ConfigOption("cluster_site_relation",
@@ -469,6 +484,25 @@ def validate_options(self):
 
                 continue
 
+            if option.key == "vm_exclude_disk_sync_by_tag":
+
+                option.set_value(quoted_split(option.value))
+
+                continue
+
+            if option.key == "vm_exclude_disk_sync":
+
+                re_compiled = None
+                try:
+                    re_compiled = re.compile(option.value)
+                except Exception as e:
+                    log.error(f"Problem parsing regular expression for '{self.source_name}.{option.key}': {e}")
+                    self.set_validation_failed()
+
+                option.set_value(re_compiled)
+
+                continue
+
             if "relation" in option.key and "vlan_group_relation" not in option.key:
 
                 relation_data = list()

module/sources/vmware/connection.py

@@ -1128,24 +1128,46 @@ def add_device_vm_to_inventory(self, object_type, object_data, pnic_data=None, v
         if version.parse(self.inventory.netbox_api_version) >= version.parse("3.7.0") and \
                 object_type == NBVM and disk_data is not None and len(disk_data) > 0:
 
-            # create pairs of existing and discovered disks.
-            # currently these disks are only used within the VM model. that's why we use this simple approach and
-            # just rewrite disk as they appear in order.
-            # otherwise we would need to implement a matching function like matching interfaces.
-            disk_zip_list = zip_longest(
-                sorted(device_vm_object.get_virtual_disks(), key=lambda x: grab(x, "data.name")),
-                sorted(disk_data, key=lambda x: x.get("name")),
-                fillvalue="X")
-
-            for existing, discovered in disk_zip_list:
-                if existing == "X":
-                    self.inventory.add_object(NBVirtualDisk, source=self,
-                                              data={**discovered, **{"virtual_machine": device_vm_object}}, )
-                elif discovered == "X":
-                    log.info(f"{existing.name} '{existing.get_display_name(including_second_key=True)}' has been deleted")
-                    existing.deleted = True
-                else:
-                    existing.update(data=discovered, source=self)
+            # Skip disk updates for VMs that match exclusion filters
+            skip_disk_sync = False
+
+            # Check if VM name matches vm_exclude_disk_sync filter
+            if hasattr(self.settings, 'vm_exclude_disk_sync') and self.settings.vm_exclude_disk_sync is not None:
+                if self.settings.vm_exclude_disk_sync.match(object_data.get("name")):
+                    log.debug(f"VM '{object_data.get('name')}' matches vm_exclude_disk_sync filter. "
+                              f"Skipping disk synchronization.")
+                    skip_disk_sync = True
+
+            # Check if VM has any tags that match vm_exclude_disk_sync_by_tag filter
+            if not skip_disk_sync and hasattr(self.settings, 'vm_exclude_disk_sync_by_tag') and \
+                    self.settings.vm_exclude_disk_sync_by_tag is not None:
+                vm_tags = [NetBoxObject.extract_tag_name(tag) for tag in device_vm_object.data.get("tags", list())]
+                for exclude_tag in self.settings.vm_exclude_disk_sync_by_tag:
+                    if exclude_tag in vm_tags:
+                        log.debug(f"VM '{object_data.get('name')}' has tag '{exclude_tag}' which matches "
+                                  f"vm_exclude_disk_sync_by_tag filter. Skipping disk synchronization.")
+                        skip_disk_sync = True
+                        break
+
+            if not skip_disk_sync:
+                # create pairs of existing and discovered disks.
+                # currently these disks are only used within the VM model. that's why we use this simple approach and
+                # just rewrite disk as they appear in order.
+                # otherwise we would need to implement a matching function like matching interfaces.
+                disk_zip_list = zip_longest(
+                    sorted(device_vm_object.get_virtual_disks(), key=lambda x: grab(x, "data.name")),
+                    sorted(disk_data, key=lambda x: x.get("name")),
+                    fillvalue="X")
+
+                for existing, discovered in disk_zip_list:
+                    if existing == "X":
+                        self.inventory.add_object(NBVirtualDisk, source=self,
+                                                  data={**discovered, **{"virtual_machine": device_vm_object}}, )
+                    elif discovered == "X":
+                        log.info(f"{existing.name} '{existing.get_display_name(including_second_key=True)}' has been deleted")
+                        existing.deleted = True
+                    else:
+                        existing.update(data=discovered, source=self)
 
         # compile all nic data into one dictionary
         if object_type == NBVM:

settings-example.ini

@@ -405,6 +405,14 @@ password = super-secret
 ; behavior also applies for VM disk sizes.
 ;vm_disk_and_ram_in_decimal = True
 
+; defines a comma separated list of VM names (regex) where disk synchronization will be excluded. 
+; A VM matching this filter will still be synced to NetBox, but its disk information won't be updated.
+;vm_exclude_disk_sync = backup-.*, temp-.*
+
+; defines a comma separated list of vCenter tags which (if assigned to a VM) will exclude this VM from 
+; disk synchronization. A VM with this tag will still be synced to NetBox, but its disk information won't be updated.
+;vm_exclude_disk_sync_by_tag = backup-vm, veeam-job
+
 [source/my-redfish-example]
 
 ; Defines if this source is enabled or not