dheeraj12347 commented on code in PR #14493:
URL: https://github.com/apache/cloudstack/pull/14493#discussion_r4591850650


##########
test/integration/smoke/test_restore_vm.py:
##########
@@ -51,11 +51,8 @@ def setUpClass(cls):
         cls.disk_offering = DiskOffering.create(cls.apiclient, 
cls.services["disk_offering"], disksize='8')
         cls._cleanup.append(cls.disk_offering)
 
-        template_t1 = Template.register(cls.apiclient, 
cls.services["test_templates"][
-            cls.hypervisor.lower() if cls.hypervisor.lower() != 'simulator' 
else 'xenserver'],
-                                        zoneid=cls.zone.id, 
hypervisor=cls.hypervisor.lower())
-        cls._cleanup.append(template_t1)
-        template_t1.download(cls.apiclient)
+        # the first template is the shared public test template, the restore 
needs another one
+        template_t1 = get_test_template(cls.apiclient, cls.zone.id, 
cls.hypervisor)

Review Comment:
   Could we add an explicit check for `FAILED` after calling 
`get_test_template()` here?
   
   `get_test_template()` can return `FAILED` if template discovery fails or no 
suitable template is found. Accessing `template_t1.id` immediately afterwards 
would then raise an `AttributeError`, rather than reporting the actual template 
retrieval failure.
   
   An explicit check would make the test setup failure clearer and consistent 
with the other updated smoke tests.



##########
setup/db/templates.sql:
##########
@@ -34,16 +34,16 @@ INSERT INTO `cloud`.`vm_template` (id, uuid, unique_name, 
name, public, created,
     VALUES (2, UUID(), 'centos53-x86_64', 'CentOS 5.3(64-bit) no GUI 
(XenServer)', 1, now(), now(), 'BUILTIN', 0, 64, 1, 
'http://download.cloudstack.org/templates/builtin/f59f18fb-ae94-4f97-afd2-f84755767aca.vhd.bz2',
 'b63d854a9560c013142567bbae8d98cf', 0, 'CentOS 5.3(64-bit) no GUI 
(XenServer)', 'VHD', 12, 1, 1, 'XenServer', 1);
 
 INSERT INTO `cloud`.`vm_template` (id, uuid, unique_name, name, public, 
created, type, hvm, bits, account_id, url, checksum, display_text, 
enable_password, format, guest_os_id, featured, cross_zones, hypervisor_type, 
extractable)
-    VALUES (4, UUID(), 'centos55-x86_64', 'CentOS 5.5(64-bit) no GUI (KVM)', 
1, now(), 'BUILTIN', 0, 64, 1, 
'http://download.cloudstack.org/releases/2.2.0/eec2209b-9875-3c8d-92be-c001bd8a0faf.qcow2.bz2',
 'ed0e788280ff2912ea40f7f91ca7a249', 'CentOS 5.5(64-bit) no GUI (KVM)', 0, 
'QCOW2', 112, 1, 1, 'KVM', 1);
+    VALUES (4, UUID(), 'debian13-x86_64-kvm', 'Debian 13 (64-bit) genericcloud 
(KVM)', 1, now(), 'BUILTIN', 0, 64, 1, 
'http://download.cloudstack.org/templates/debian-13-genericcloud-amd64/debian-13-genericcloud-amd64-20261026-2643-acs.qcow2',
 
'{SHA-512}bd95d28cc1e42e2f9d54c16c1015fee0204677dc6122c9a10bec8aaf0b3730b426f47e1c50e046f21add95f917344fe07089d791eb996fbd996257879022aafe',
 'Debian 13 (64-bit) genericcloud (KVM)', 0, 'QCOW2', 184, 1, 1, 'KVM', 1);

Review Comment:
   > [@dheeraj12347](https://github.com/dheeraj12347) the guest os will be 
later changed to "Debian GNU/Linux 12 (64-bit)" , you may have missed the 
changes in engine/schema/src/main/resources/META-INF/db/schema-42300to2400.sql
   > 
   > "Debian 13" is not officially supported in CloudStack. It is too new to 
some old VMware/Xen/XCPng versions which does not support "Debian 13". For 
example on VMware:
   > 
   > <img alt="image" width="2000" height="1005" 
src="https://private-user-images.githubusercontent.com/57355700/704943041-ba260b1d-f0df-4390-83c3-fcb3acccf030.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.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.bQEdikEbyiXMeWae63G_A1Kp8ndWCk3zvePK5tTWQng";>
   > Actually "Debian 12" is not supported by some old VMware versions
   > 
   > <img alt="image" width="2000" height="1071" 
src="https://private-user-images.githubusercontent.com/57355700/704944518-51fdd614-9523-4408-9d5c-3f4c8d9601fb.png?jwt=eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3OTQ4NjQyMTAsIm5iZiI6MTc5NDg2MzkxMCwicGF0aCI6Ii81NzM1NTcwMC83MDQ5NDQ1MTgtNTFmZGQ2MTQtOTUyMy00NDA4LTlkNWMtM2Y0YzhkOTYwMWZiLnBuZz9YLUFtei1BbGdvcml0aG09QVdTNC1ITUFDLVNIQTI1NiZYLUFtei1DcmVkZW50aWFsPUFLSUFWQ09EWUxTQTUzUFFLNFpBJTJGMjAyNjExMTYlMkZ1cy1lYXN0LTElMkZzMyUyRmF3czRfcmVxdWVzdCZYLUFtei1EYXRlPTIwMjYxMTE2VDIxMTgzMFomWC1BbXotRXhwaXJlcz0zMDAmWC1BbXotU2lnbmF0dXJlPTEwMWVlOTg4MzFmMmU5MTRiZGNmYzJkMzIzNTI4MzRhZmZlMmRhMTJhNTA4MjZlMzljMjY4ZDUyZTEyYzA3NmUmWC1BbXotU2lnbmVkSGVhZGVycz1ob3N0JnJlc3BvbnNlLWNvbnRlbnQtdHlwZT1pbWFnZSUyRnBuZyJ9.tMjgptHxBjOA4kErHe_IrNrGTOQJShe-5ghEsAIK_cE";>
   
   Thanks for clarifying the migration.
   
   I understand that the Debian 13 image is intentionally associated with 
Debian 12 guest OS metadata for compatibility. I think the important point is 
to verify the supported guest OS mappings across the target VMware, XenServer 
and XCP-ng versions.
   
   Would it make sense to document the minimum supported hypervisor versions 
for these built-in templates, or handle guest OS mappings per hypervisor 
version where necessary? I would be cautious about switching all templates to 
Debian 11 without validating the Debian 13 image and its guest tools on those 
older versions.



##########
test/integration/smoke/test_scale_vm.py:
##########
@@ -249,18 +247,7 @@ def test_01_scale_vm(self):
             self.apiclient,
             isdynamicallyscalable='true')
 
-        if self.is_host_xcpng8(res):
-            self.debug("Only scaling for CPU for XCP-ng 8")
-            offering_data = self.services["service_offerings"]["big"]
-            offering_data["cpunumber"] = 2
-            offering_data["memory"] = self.virtual_machine.memory
-            self.bigger_offering = ServiceOffering.create(
-                self.apiclient,
-                offering_data
-            )
-            self.cleanup.append(self.bigger_offering)
-        else:
-            self.bigger_offering = self.big_offering

Review Comment:
   Could we confirm that memory scaling is supported on XCP-ng 8.x with the 
Debian 13 built-in template?
   
   The previous implementation had a specific XCP-ng 8 path that kept the VM's 
memory unchanged and only scaled CPU. This change appears to remove that 
special handling and use the larger service offering for all hypervisors.
   
   Given the scale-VM failures in the XCP-ng CI runs, could you clarify whether 
memory hot-add has been verified on XCP-ng 8.2/8.3? If it is still unsupported, 
should we retain the CPU-only scaling path for those versions?



-- 
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: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to