How to call scaleVirtualMachine api with details parameter

[YashasAnand]

I am trying to call scaleVirtualMachine API using python, but the documentation is very vague on how to use details parameter
https://cloudstack.apache.org/api/apidocs-4.19/apis/scaleVirtualMachine.html

I am trying to create request like this, is this wrong? can anyone pls suggest what is right way to call this api?

request = {
'command': 'scaleVirtualMachine',
'id': 'xys', # ID of the system VM
'serviceofferingid': 'xyz',
'response': 'json',
'details' : [
{"name": "memory", "value": "3072"},
{"name": "cpunumber", "value": "2"}
]
}

I get error

{
"scalevirtualmachineresponse": {
"uuidList": [],
"errorcode": 431,
"cserrorcode": 9999,
"errortext": "Unable to execute API command scalevirtualmachine due to invalid value [{"name": "memory", "value": "3072"}, {"name": "cpunumber", "value": "2"}] for parameter details"
}
}

[weizhouapache]

Can you try

["memory": "3072", "cpunumber": "2"]

or similar ?

[bernardodemarco]

Hello, @YashasAnand

The details parameter of the scaleVirtualMachine API is a MAP. Basically, the MAP parameter type in CloudStack expects the keys and values as:

details[0].key=value

So, for example, a VM with 512 MB of RAM and 1 vCPU can be scaled to 1024 MB of RAM and 2 vCPUs by executing the following API call through CloudMonkey:

scale virtualmachine id=<vm-id> serviceofferingid=<compute-offering-id> details[0].memory=1024 details[0].cpuNumber=2

How does your Python script sends the request? Does each key in the request dict represent a query param in the HTTP request? If so, then one solution would be to define the details parameter to be something like:

request = {
    'command': 'scaleVirtualMachine',
    'id': '<virtual-machine-id>',
    'serviceofferingid': '<compute-offering-id>',
    'response': 'json',
    'details[0].memory': 1024 ,
    'details[0].cpuNumber': 2
}

[YashasAnand]

Hi @bernardodemarco @weizhouapache Thank you for your respective responses. Below is my python code.

def generate_signature(baseurl, request, secretkey):

   encoded_request = {urllib.parse.quote_plus(k): urllib.parse.quote_plus(str(v)) for k, v in request.items()} 
   sorted_request = sorted(encoded_request.items(), key=lambda x: x[0].lower())
   sig_str = '&'.join(['='.join([k.lower(), v.lower()]) for k, v in sorted_request])

   sig = hmac.new(secretkey.encode('utf-8'), sig_str.encode('utf-8'), hashlib.sha1).digest()

   encoded_sig = base64.b64encode(sig).decode('utf-8')
   final_signature = urllib.parse.quote_plus(encoded_sig)

   request_str = '&'.join(['='.join([k, v]) for k, v in encoded_request.items()])
   return final_signature, request_str

baseurl = 'basrurl'

request = {
'command': 'scaleVirtualMachine',  
'id': 'id',  # ID of the system VM
'serviceofferingid': 'serviceofferingid', 
'response': 'json', 
'details[0].memory': 3072,
'details[0].cpuNumber': 2
}
secretkey = 'secret key'
signature, request_str = generate_signature(baseurl, request, secretkey)
req = baseurl + request_str + '&signature=' + signature
print("Final Request URL:", req)

I am getting response like this but i dont observe any scaling happening on the vm even though i have domain admin access. Pls can you suggest what might be the issue.

{ "scalevirtualmachineresponse": { "jobid": "f75d1e48-132e-4b5c-9241-9115271dc08e" } }

[bernardodemarco]

APIs in CloudStack can be synchronous/blocking, or asynchronous/non-blocking. Async APIs return a jobid, which can be used to poll the result of the API call.

Since the scaleVirtualMachine API is asynchronous, the jobid was returned, as expected. It means that the request was successfully authenticated and its response can be retrieved using the queryAsyncJobResult API:

query asyncjobresult jobid="<job-id>"

When retrieving the result of a job related to the scaleVirtualMachine API, a successful response will be something like:

(localcloud) 🐱 > query asyncjobresult jobid=75f6ebd3-e4ed-477c-b6ff-2577e2e56c30
{
  (...)
  "jobresult": {
    "virtualmachine": {
      "account": "admin",
      "affinitygroup": [],
      "cpunumber": 1,
      "cpuspeed": 512,
      "created": "2024-11-21T15:07:09+0000",
      "deleteprotection": false,
      "details": {
        "Message.ReservedCapacityFreed.Flag": "true",
        "cpuNumber": "1",
        "cpuSpeed": "512",
        "memory": "512"
      },
      (...)
    }
  },
  "jobresultcode": 0,
  "jobresulttype": "object",
  "jobstatus": 1,
  "userid": "203addbc-4b71-11ef-87b4-cec422422af1"
}

Given that, could you check the return of the queryAsyncJobResult API?