IT Toolbox

VMware Center / Virtualization

VMware / Virtualisation

VMware / Virtualization Center

Guided diagnostics and field commands for ESXi, vCenter, VMs, snapshots, datastores, shared storage, virtual networking, HA, DRS and backup.

InventaireLister les VM sur un hôte
vim-cmd vmsvc/getallvms
InventaireLister les processus VM
esxcli vm process list
InventaireShow les datastores
esxcli storage filesystem list
VMÉtat d’alimentation d’une VM
vim-cmd vmsvc/power.getstate <VMID>
VMDémarrer une VM
vim-cmd vmsvc/power.on <VMID>
VMÉteindre proprement une VM
vim-cmd vmsvc/power.shutdown <VMID>
VMArrêt forcé d’une VM
vim-cmd vmsvc/power.off <VMID>
SnapshotLister les snapshots d’une VM
vim-cmd vmsvc/snapshot.get <VMID>
SnapshotDelete tous les snapshots
vim-cmd vmsvc/snapshot.removeall <VMID>
StorageLister les devices
esxcli storage core device list
StorageLister les paths
esxcli storage core path list
StorageLister les politiques NMP
esxcli storage nmp device list
StorageLister les adaptateurs iSCSI
esxcli iscsi adapter list
StorageDiagnostic lock VMDK
vmkfstools -D /vmfs/volumes/<DATASTORE>/<VM>/<DISQUE>.vmdk
NetworkingLister les NIC physiques
esxcli network nic list
NetworkingLister les interfaces vmkernel
esxcli network ip interface list
NetworkingIPv4 vmkernel
esxcli network ip interface ipv4 get
NetworkingRoutes IPv4
esxcli network ip route ipv4 list
NetworkingLister les vSwitch Standard
esxcli network vswitch standard list
NetworkingTester depuis un vmkernel
vmkping -I vmk0 <IP_CIBLE>
ServicesÉtat hostd
/etc/init.d/hostd status
ServicesÉtat vpxa
/etc/init.d/vpxa status
ServicesRedémarrer agents management
services.sh restart
HAÉtat agent FDM
/etc/init.d/fdm status
HADerniers logs FDM
tail -n 100 /var/log/fdm.log
LogsDerniers logs VM
tail -n 100 /vmfs/volumes/<DATASTORE>/<VM>/vmware.log
LogsKernel ESXi
tail -n 100 /var/log/vmkernel.log
LogsHostd
tail -n 100 /var/log/hostd.log
PowerCLIConnexion vCenter
Connect-VIServer -Server <VCENTER>
PowerCLILister les VM
Get-VM
PowerCLILister les snapshots
Get-VM | Get-Snapshot
PowerCLICapacité des datastores
Get-Datastore | Select Name,CapacityGB,FreeSpaceGB
PowerCLILister les hôtes
Get-VMHost
PowerCLIÉtat VMware Tools
Get-VM | Select Name,@{N='ToolsStatus';E={$_.ExtensionData.Guest.ToolsStatus}}
PowerCLILister réseaux de VM
Get-VM | Get-NetworkAdapter | Select Parent,Name,NetworkName,MacAddress

Calculateur de capacité datastore

Estimez rapidement l’espace restant après ajout ou croissance de VM.

1

Snapshot ≠ backup

Un snapshot est temporaire et dépend du datastore et de la VM d’origine.

2

Éviter les snapshots longs

Les deltas peuvent grossir et compliquer consolidation, backup et performances.

3

Ne jamais supprimer un VMDK à l’aveugle

Identifiez la chaîne de disques et les descriptors avant toute suppression.

4

Conserver de la marge

Un datastore presque plein peut empêcher snapshot, consolidation ou démarrage de VM.

5

Tester le vmkernel

Pour vMotion, iSCSI ou management, testez depuis le vmkernel réellement concerné.

6

Documenter les paths SAN

Conservez HBA, WWPN, zoning, LUN et politiques multipath dans la documentation d’infrastructure.

Important: les commandes et services VMware peuvent varier selon les versions ESXi/vCenter. Les actions d’arrêt forcé, snapshot ou manipulation de VMDK doivent être réalisées avec prudence.
Field troubleshooting

Center mission

8 playbooks

Troubleshoot ESXi/vCenter by separating host capacity, datastore, VM state, snapshots, networking, VMware Tools, vMotion and HA before infrastructure changes.

Quick triage

  • Check vCenter/ESXi alarms and triggering event.
  • Check host health, CPU/RAM and datastores.
  • Identify whether one VM, one datastore or the whole cluster is affected.
  • Check snapshots/consolidation before storage operations.
  • Keep tasks/events and timestamps before restarting agent/host.

Decision tree

One VM only→

Start with VM state, files, snapshot, Tools and virtual network.

Several VMs on same datastore→

Prioritize capacity, latency and storage connectivity.

Whole host affected→

Check management, resources, agents and hardware.

Cluster/HA affected→

Check vCenter, HA heartbeat, admission control and management network.

Intervention playbooks

Start read-only, collect evidence, then change one variable at a time.

01VM will not power onControlled change
Symptom

Power On fails with resource, file lock or configuration error.

Checks

  • Read exact task/event in vCenter.
  • Check datastore free space, host resources and VM files.
  • Check lock or snapshot/consolidation.

Commands / evidence

vim-cmd vmsvc/getallvmsvim-cmd vmsvc/power.getstate <vmid>esxcli storage filesystem list

Expected result

VM has resources, accessible files and no illegitimate lock.

Corrective actions

  • Fix identified resource/datastore/lock before re-register or heavier action.
  • Avoid manually deleting snapshot files.

Escalate when

VM files corrupted, multi-host lock or unstable storage.

02Datastore nearly fullIntrusive / escalation
Symptom

Capacity alarm, snapshot failures or VMs suspended due to space.

Checks

  • Measure capacity/free and recent growth.
  • Identify snapshots, ISOs and orphan files without blind deletion.
  • Check array-side thin provisioning.

Commands / evidence

esxcli storage filesystem listdu -h /vmfs/volumes/<datastore>/<vm> 2>/dev/nullvim-cmd vmsvc/snapshot.get <vmid>

Expected result

Growth source is identified and safety margin is restored.

Corrective actions

  • Consolidate/remove snapshots through vSphere with enough working space.
  • Extend or migrate if capacity use is legitimate.

Escalate when

Critical datastore >95%, array also full or consolidation needs substantial temporary space.

03Snapshot / consolidation neededIntrusive / escalation
Symptom

Consolidation needed, very old snapshot or deletion fails.

Checks

  • List snapshots and age/size.
  • Check sufficient datastore space.
  • Check backup tasks that may hold snapshot.

Commands / evidence

vim-cmd vmsvc/snapshot.get <vmid>esxcli storage filesystem listls -lh /vmfs/volumes/<datastore>/<vm>/

Expected result

Snapshot chain is understood, datastore has margin and no competing task.

Corrective actions

  • Run consolidation from vSphere in a monitored window.
  • Stop cause of recurring snapshots on backup side.

Escalate when

Inconsistent chain, huge delta or consolidation fails with disk/lock error.

04VM has no networkRead-only
Symptom

VM is running but has no network, gateway or services.

Checks

  • Check vNIC is connected and correct port group.
  • Check port group/vSwitch/dvSwitch VLAN.
  • Compare another VM on same port group.

Commands / evidence

esxcli network vm listesxcli network nic listesxcli network vswitch standard list

Expected result

vNIC is connected to correct network and physical uplink is up.

Corrective actions

  • Fix port group/VLAN/uplink according to failing layer.
  • Do not change multiple trunk VLANs without impact analysis.

Escalate when

dvSwitch/vCenter inconsistency, complex teaming or shared physical network outage.

05ESXi host overloadedRead-only
Symptom

VMs slow, high ready time, balloon/swap memory or datastore latency.

Checks

  • Separate CPU contention, memory pressure and storage latency.
  • Identify consuming VMs.
  • Compare other cluster hosts.

Commands / evidence

esxtopesxcli hardware memory getesxcli system stats uptime get

Expected result

Limiting resource is identified, not just a global average.

Corrective actions

  • Rebalance/migrate load or fix responsible VM.
  • Avoid further overcommit to hide contention.

Escalate when

Cluster-wide contention, slow shared storage or capacity planning needed.

06VMware Tools missing or outdatedControlled change
Symptom

Guest status unknown, graceful shutdown unavailable or old network/storage drivers.

Checks

  • Read Tools state in vCenter.
  • Check OS version and compatibility.
  • Plan reboot if drivers require it.

Commands / evidence

vim-cmd vmsvc/get.guest <vmid>vmware-toolbox-cmd -vsystemctl status open-vm-tools

Expected result

Tools runs and vCenter receives heartbeat/guest info.

Corrective actions

  • Update using supported method for OS.
  • Test snapshot backup afterward if VSS/quiescing is involved.

Escalate when

Legacy OS, critical driver or Tools update causes regression.

07vMotion failsRead-only
Symptom

Live migration fails at precheck, network or storage stage.

Checks

  • Read compatibility/precheck error.
  • Check vMotion vmkernel, MTU and connectivity.
  • Check shared datastore access and EVC/CPU.

Commands / evidence

esxcli network ip interface listvmkping -I vmk<Vmotion> <peer-ip>esxcli network ip route ipv4 list

Expected result

vmkernel interfaces reach each other with expected MTU and hosts are compatible.

Corrective actions

  • Fix targeted vmkernel network or compatibility issue.
  • Retest a non-critical VM before bulk migration.

Escalate when

Complex vDS, NSX, EVC or multipath storage.

08HA / host isolationIntrusive / escalation
Symptom

Host disconnected, isolation response or VMs restarted by HA.

Checks

  • Check management network and isolation gateways.
  • Check HA agents and datastore heartbeat access.
  • Read cluster events before reconnecting anything.

Commands / evidence

vmkping <management-gateway>esxcli network ip interface list/etc/init.d/hostd status

Expected result

Host remains management-reachable and HA sees enough heartbeats.

Corrective actions

  • Fix management network before HA reconfiguration.
  • Validate cluster capacity before maintenance.

Escalate when

Network/storage split-brain, multiple isolated hosts or double-start VM risk.

End-of-intervention checklist

  • Verify vCenter alarms returned green.
  • Check datastore, host and VM after fix.
  • Retest network/Tools/application from affected VM.
  • Document snapshot, migration or infrastructure change.
  • Monitor backup/HA tasks after intervention.

Continue in BAOI

Related cheat sheetVMware — operations IT toolsCalculate, inspect or generate without leaving the workflow. ProceduresFollow a controlled implementation procedure. Known failuresCross-check the symptom with known failure patterns.
♡ 0