From 58db5370936e9acdda1cb328fc7cc73cf2bced6b Mon Sep 17 00:00:00 2001
From: "Joshua M. Boniface" <joshua@boniface.me>
Date: Mon, 13 Sep 2021 01:46:53 -0400
Subject: [PATCH] Add memory and vCPU checks to VM define/modify

Ensures that a VM won't:

(a) Have provisioned more RAM than there is available on a given node.
Due to memory overprovisioning, this is simply a "is the VM memory count
more than the node count", and doesn't factor in free or used memory on
a node, total cluster usage, etc. So if a node has 64GB total RAM, the
VM limit is 64GB. It is up to an administrator to ensure sanity *below*
that value.

(b) Have provisioned more vCPUs than there are CPU cores on the node,
minus 2 to account for hypervisor/storage processes. Will ensure there
is no severe CPU contention caused by a single VM having more vCPUs than
there are actual execution threads available.

Closes #139
---
 daemon-common/vm.py | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/daemon-common/vm.py b/daemon-common/vm.py
index 5b6cf889..56ce03dd 100644
--- a/daemon-common/vm.py
+++ b/daemon-common/vm.py
@@ -196,6 +196,16 @@ def define_vm(zkhandler, config_data, target_node, node_limit, node_selector, no
         if not valid_node:
             return False, 'ERROR: Specified node "{}" is invalid.'.format(target_node)
 
+    # Validate the new RAM against the current active node
+    node_total_memory = int(zkhandler.read(('node.memory.total', target_node)))
+    if int(parsed_xml.memory.text) >= node_total_memory:
+        return False, 'ERROR: VM configuration specifies more memory ({} MiB) than node "{}" has available ({} MiB).'.format(parsed_xml.memory.text, target_node, node_total_memory)
+
+    # Validate the number of vCPUs against the current active node
+    node_total_cpus = int(zkhandler.read(('node.data.static', target_node)).split()[0])
+    if (node_total_cpus - 2) <= int(parsed_xml.vcpu.text):
+        return False, 'ERROR: VM configuration specifies more vCPUs ({}) than node "{}" has available ({} minus 2).'.format(parsed_xml.vcpu.text, target_node, node_total_cpus)
+
     # If a SR-IOV network device is being added, set its used state
     dnetworks = common.getDomainNetworks(parsed_xml, {})
     for network in dnetworks:
@@ -429,6 +439,17 @@ def modify_vm(zkhandler, domain, restart, new_vm_config):
     old_parsed_xml = lxml.objectify.fromstring(old_vm_config)
     old_dnetworks = common.getDomainNetworks(old_parsed_xml, {})
 
+    # Validate the new RAM against the current active node
+    node_name = zkhandler.read(('domain.node', dom_uuid))
+    node_total_memory = int(zkhandler.read(('node.memory.total', node_name)))
+    if int(parsed_xml.memory.text) >= node_total_memory:
+        return False, 'ERROR: Updated VM configuration specifies more memory ({} MiB) than node "{}" has available ({} MiB).'.format(parsed_xml.memory.text, node_name, node_total_memory)
+
+    # Validate the number of vCPUs against the current active node
+    node_total_cpus = int(zkhandler.read(('node.data.static', node_name)).split()[0])
+    if (node_total_cpus - 2) <= int(parsed_xml.vcpu.text):
+        return False, 'ERROR: Updated VM configuration specifies more vCPUs ({}) than node "{}" has available ({} minus 2).'.format(parsed_xml.vcpu.text, node_name, node_total_cpus)
+
     # If a SR-IOV network device is being added, set its used state
     dnetworks = common.getDomainNetworks(parsed_xml, {})
     for network in dnetworks: