devops-ai-toolkit/src/devops_ai_toolkit/knowledge/data/vmware.yaml at main · devopsaitoolkit/devops-ai-toolkit · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
- id: vmware.insufficient_resources_ha
  technology: vmware
  title: "vSphere: 'Insufficient resources to satisfy configured failover'"
  summary: >-
    vSphere HA cannot guarantee failover capacity, or a VM cannot power on
    because the cluster/host lacks the CPU or memory the admission control needs.
  applies_to: [log, command_output, error_string]
  match:
    any_of:
      - "Insufficient resources to satisfy (configured failover|HA)"
      - "Insufficient resources to satisfy vSphere HA"
      - "failed to power on .* insufficient"
    weight: 0.72
  root_causes:
    - title: "HA admission control reserves more capacity than is free"
      description: >-
        The cluster's failover capacity policy reserves resources that the
        current load no longer leaves available, blocking power-on.
      confidence: 0.6
      category: capacity
    - title: "Host failures reduced usable capacity"
      description: "One or more hosts are down/maintenance, shrinking cluster capacity."
      confidence: 0.5
      category: availability
    - title: "Oversized VM reservations"
      description: "A VM's CPU/memory reservation cannot be satisfied by any single host."
      confidence: 0.4
      category: configuration
  diagnostic_commands:
    - command: "esxcli system maintenanceMode get"
      explanation: "Checks whether the host is in maintenance mode (read-only)."
      expected_output: "'Disabled' when the host should be serving workloads."
    - command: "esxcli vm process list"
      explanation: "Lists running VMs and their world IDs on the host."
      expected_output: "The set of powered-on VMs consuming capacity."
  suggested_fixes:
    - title: "Free capacity or relax admission control"
      description: >-
        Bring hosts out of maintenance, migrate or right-size VM reservations,
        or adjust the HA admission control policy to match real capacity.
  references:
    - title: "vSphere Availability (HA) documentation"
      url: "https://docs.vmware.com/en/VMware-vSphere/index.html"
      source: "official docs"
  warnings:
    - message: "Relaxing HA admission control reduces guaranteed failover capacity."
      severity: medium
  best_practices:
    - "Size clusters with enough headroom for at least one host failure."
  prevention:
    - "Review HA admission control after adding or removing hosts."
  tags: [vmware, vsphere, ha, capacity]

- id: vmware.cannot_connect_esxi_host
  technology: vmware
  title: "vCenter: host 'Not Responding' / cannot connect to ESXi"
  summary: >-
    vCenter shows an ESXi host as Not Responding or Disconnected and management
    operations on it fail.
  applies_to: [log, command_output, error_string]
  match:
    any_of:
      - "Not Responding"
      - "host .* is disconnected"
      - "Cannot (contact|connect to) .* host"
      - "vpxa"
    weight: 0.68
  root_causes:
    - title: "Management agents (hostd/vpxa) are unresponsive"
      description: "The ESXi management agents have hung and need a restart."
      confidence: 0.55
      category: service
    - title: "Network partition between vCenter and the host"
      description: "Management network or firewall changes broke connectivity (port 902/443)."
      confidence: 0.5
      category: network
    - title: "Host resource exhaustion"
      description: "The host ran out of memory/CPU for management, dropping the connection."
      confidence: 0.35
      category: capacity
  diagnostic_commands:
    - command: "ping <esxi-mgmt-ip>"
      explanation: "Confirms basic network reachability of the host management interface."
      expected_output: "Replies when the management network is intact."
    - command: "esxcli network ip connection list | grep -E '902|443'"
      explanation: "Checks the management ports vCenter uses to reach the host (read-only)."
      expected_output: "ESTABLISHED/LISTEN entries on 902/443."
  suggested_fixes:
    - title: "Restore host management connectivity"
      description: >-
        Verify the management network and firewall, then restart the ESXi
        management agents if they are hung (services.sh restart) from the host console.
  references:
    - title: "VMware ESXi troubleshooting documentation"
      url: "https://docs.vmware.com/en/VMware-vSphere/index.html"
      source: "official docs"
  best_practices:
    - "Redundant management NICs reduce host-isolation events."
  prevention:
    - "Monitor hostd/vpxa health and management-network reachability."
  tags: [vmware, vsphere, esxi, vcenter, network]