About this article: This article is reviewed against the cited sources. This is source research, not lab validation. Check the stated versions and your own requirements before making changes.
The message "Another task is already in progress." means an earlier task on that virtual machine, or on the host, has not finished as far as vCenter or the ESXi host is concerned. Broadcom KB 314602 explains that the vCenter Server view and the host's view of tasks can differ, and that the error may appear after a task in vCenter times out. Before you retry, restart anything or reboot a host, find the task on the host, read its state, and decide whether it is working or stale.
What the error means
KB 314602, "Collecting information about tasks in VMware ESXi" (Updated On: 07-27-2026, ESXi 6.x, 7.x and 8.x), says that "there may be differences between what vCenter Server and an ESXi host considers tasks". The error can appear when a task within vCenter Server times out and you then try another task.
Broadcom KB 310967 (Updated On: 09-10-2026) explains the timeout: vCenter Server has a default 15-minute timeout for any task, and a task that vCenter reports as timed out may continue to run on the ESXi host. It also says certain tasks, such as a snapshot consolidation, "may take a long time to complete and should not be interrupted". That KB lists a related message: "Operation failed since another task is in progress."
Similar wording appears in other situations. KB 344709 (Updated On: 06-17-2026) lists "Another task in progress" with greyed-out VM menus, and gives running tasks, file locks, a backup in progress, permissions and a damaged .vmx as causes. KB 454728 (Updated On: 09-02-2026) shows the same message when a ScanV2 task failed with TaskInProgress because a remediation task (InstallV2) had stalled on ESX 8.0 U3; the KB also names high storage latency and UNMAP-driven ATS lock contention on the datastore. Stuck snapshot work is covered in more depth in "A snapshot removal or consolidation task is stuck, or fails with a lock error".
Likely causes
| Cause | What you see | How to confirm |
|---|---|---|
| A long task still runs on the host after vCenter timed it out | vCenter shows the task failed or timed out, yet a retry gives the error | vim-cmd vmsvc/get.tasklist lists a task, and task_info shows state = "running" |
| Backup software starts the next snapshot removal while the first is running | Repeated failures at regular intervals | KB 392609 found a backup tool initiating snapshot removal every 5 minutes; compare the backup schedule |
| A stale task that cannot be cancelled | The task list holds a task whose progress does not change | Two task_info reads a few minutes apart show the same progress |
| Stale change-tracking files or file locks block a power-on (ESXi 8.x) | Power-on hangs at 33 or 66 percent | The power-on task stalls at 33 or 66 percent (KB 440973); the KB's procedure checks the VMDK locks |
| Host patching in progress | The error on a host scan or remediation | Active InstallV2 task (KB 454728) |
Permissions or a damaged .vmx |
Greyed-out VM menus | KB 344709 steps |
Prove the cause before changing anything
Connect to the host that runs the VM with SSH or its console. These commands only read.
1. Find the VM's ID.
vim-cmd vmsvc/getallvms
The ID is in the Vmid column.
2. List the tasks for that VM.
vim-cmd vmsvc/get.tasklist VMID
Replace VMID with the number from step 1. KB 344709 says that with no task running the output is (ManagedObjectReference) [], and that if a task is running the output lists it, for example vim.Task:haTask-8-vim.VirtualMachine.createSnapshot-534613324. The part after vim.Task: is the task identifier.
3. List all tasks on the host, to catch host-level tasks:
vim-cmd vimsvc/task_list
4. Read one task.
vim-cmd vimsvc/task_info TASK_IDENTIFIER
KB 314602 shows the output fields: name, entityName, state (the sample is running), cancelled, cancelable, progress (the sample is 15), queueTime, startTime and completeTime (<unset> while unfinished). A good result is a task whose state is running and whose progress has moved since your last read: let it finish. A bad result is a task whose progress is unchanged across several minutes and which is not cancelable.
5. For a consolidation, KB 310967 says to watch the VM's .vmdk file: if it is updated every minute, the delta files are being committed, even though the client timed out.
6. Read the backup schedule. KB 392609 attributed one case to backup software retriggering removal. Check job times against the task's startTime.
Fix
A real task is still running
Wait. KB 392609's resolution begins "Wait for the consolidation task to either complete or fail, before triggering another job to remove the snapshot." Do not retry from vCenter while the host task runs. If the backup job failed on a timeout, KB 310967 says to check the VM for a backup snapshot left behind.
KB 310967 also describes changing vCenter and host timeout values for work that takes longer than 15 minutes, such as large snapshot deletions or migrations, and says temporary changes should go back to the default afterwards. The change involves editing configuration files on the appliance and the host, and restarting vpxd. Disruptive step: these file edits and the vpxd restart change live systems, so make them in a quiet period. Follow the KB itself, with a copy of each file.
Overlapping backup jobs
Stop the next job from starting until the first finishes, and lengthen the interval or add a check in the backup tool. KB 392609 names a tool retrying every 5 minutes as the cause of its case.
A stale task that will not clear
Do this only if step 5 above shows no .vmdk update for a consolidation, and no other task, backup job or migration is using the host. KB 310967 says a consolidation "may take a long time to complete and should not be interrupted".
KB 392609 says that if the stale task cannot be killed or cancelled, restarting the ESXi host management agents can clear it, and that if that fails the host can be rebooted after placing it in maintenance mode.
In an HA cluster, Broadcom KB 316576 ("Troubleshooting issues resulting from locked virtual disks", Updated On: 04-30-2026) says, in its own step to restart the management agents: "Note: Deactivate HA Host Monitoring first to prevent an unwanted VM failover."
Disruptive step: the KB says restarting management agents may cause temporary disruptions, can lead to instability that needs a host reboot, and can affect running tasks and guest performance. In an HA cluster, deactivate HA Host Monitoring first, as KB 316576 says. Save first: note the current HA Host Monitoring setting in the cluster's vSphere Availability settings, so you can put it back, and copy the output of vim-cmd vimsvc/task_list and task_info for the stuck task into a text file kept off the host, and check the vSphere Client for other work on the host such as a migration. Back out: there is none for a restart, but turn HA Host Monitoring back on to the noted setting once the host shows Connected; allow 1 to 2 minutes for the agents to start, check that the host shows Connected in the vSphere Client, and read /var/log/hostd.log if a service does not start (KB 320280).
KB 320280, "Restarting Management Agents in ESXi" (Updated On: 10-27-2026), gives the restart commands and warnings. It says not to use services.sh restart (or restart management agents from the DCUI) on hosts that use vSAN, LACP, NSX or shared graphics. Its line "Only restart single services." follows the shared-graphics check, and it gives these single-service restarts:
/etc/init.d/hostd restart
/etc/init.d/vpxa restart
Disruptive step: a host reboot stops everything on the host. KB 392609 says to reboot the host by placing it in maintenance mode; moving the VMs off first is advice that needs DRS or manual migration, and maintenance mode can itself stall while a task is stuck on the host. Save first: record which VMs run on the host and their power state. Back out: none; power on any VM that did not restart.
Stale change-tracking files or locks
KB 440973 (ESXi 8.x) gives a procedure that checks file locks with vmfsfilelockinfo, moves change-tracking files, copies the .vmx file, edits it and reloads the VM. That is a change to the VM's files.
Risk: the KB moves the *-ctk.vmdk, *.vmsd and *.vmsn files (change-tracking files, the snapshot database and the snapshot state files) into a temp folder and sets ctkEnabled and scsi#:#.ctkEnabled to "false" in the .vmx file. That affects backup tools that depend on change tracking, and a wrong edit of the .vmx file can stop the VM from starting. Save first: the KB's own step copies the configuration file (cp vm-name.vmx vm-name.vmx.bak), and you should also list the files in the VM folder before anything is moved. Follow the KB in full. The KB does not say whether the VM must be powered off; as advice, keep it powered off while the files are moved. Back out: this holds only until the VM is powered on again, because the power-on creates new change-tracking files that moved-back files can collide with. Copy vm-name.vmx.bak back over vm-name.vmx, move the files back from the temp folder the KB creates, and reload the VM configuration with the KB's vim-cmd vmsvc/reload step. The KB gives no step to turn change tracking back on, so tell the owner of the backup tool.
Host patching
KB 454728 resolves its case by aligning remediation windows away from heavy backup jobs and by suppressing scheduled compliance scans during remediation.
Example (illustrative): A backup tool starts a snapshot on web01 at 01:00 and vCenter reports the task as timed out at 01:15. At 01:20 an administrator tries to remove the snapshot and gets "Another task is already in progress." On esx01, vim-cmd vmsvc/getallvms shows web01 with Vmid 112. vim-cmd vmsvc/get.tasklist 112 lists one task. task_info shows state = "running" with progress 15 at 01:22 and a higher figure at 01:30. The administrator waits, the task completes at 01:41, and the second request then succeeds.
Confirm the fix
vim-cmd vmsvc/get.tasklist VMIDreturns(ManagedObjectReference) [].- The operation that failed now works from vCenter.
- The VM shows no leftover snapshot or "Consolidation is needed" alert; if it does, handle that as its own fault.
- The backup schedule no longer overlaps.
What to watch afterwards
- Alert on tasks that run longer than your usual window.
- Keep a record of the longest normal snapshot removal for each large VM, so that a long task is recognised as normal.
- After any management agent restart, check that the host stays Connected.
Read next
Related articles you may want next.
- TroubleshootingA snapshot removal or consolidation task is stuck, or fails with a lock errorFind which host, process or backup proxy holds the disk file behind a stuck snapshot removal or consolidation, prove it, and clear it through scoped recovery decisions.Read article
- TroubleshootingESXi Host Not Responding or Disconnected in vCenter: Narrow the CauseESXi host Not Responding or Disconnected? Check the UDP 902 heartbeat, managed IP and hostd/vpxa before restarting agents or the host.Read article
- TroubleshootingMy VM freezes for a while when a snapshot is created or removedLearn what a VM stun is, read stun times from vmware.log, find what makes snapshot pauses long, and assess version-specific change options and their application impact.Read article
Risks and limits
- These commands have not been run or lab-tested.
- KB 314602 shows
vim-cmdoutput from one host; field order and values differ by version and task. - Reading
task_infotwice to see whether progress moves is a practical test that Broadcom's KBs do not define. Do not cancel or restart because of one reading. - KB 314602, KB 392609 and KB 344709 do not give a command to cancel a host task.
- KB 310967 documents timeout edits for vpxd and vpxa; follow it with a copy of each file.
Sources
- Collecting information about tasks in VMware ESXi — Broadcom Knowledge Base 314602, Updated On: 07-27-2026, accessed 20 November 2026
- Virtual Machine snapshot removal task is stuck with Error: 'Another task is already in progress' in vCenter server — Broadcom Knowledge Base 392609, Updated On: 09-28-2026, accessed 20 November 2026
- vCenter operation times out with the error: Operation failed since another task is in progress — Broadcom Knowledge Base 310967, Updated On: 09-10-2026, accessed 20 November 2026
- Restarting Management Agents in ESXi — Broadcom Knowledge Base 320280, Updated On: 10-27-2026, accessed 20 November 2026
- Error "Another task is already in progress" during VM power on due to stale CBT files in VMware vCenter Server — Broadcom Knowledge Base 440973, Updated On: 09-16-2026, accessed 20 November 2026
- vCenter reported "Another task is already in progress" while the host was undergoing active VIB installation — Broadcom Knowledge Base 454728, Updated On: 09-02-2026, accessed 20 November 2026
- Virtual machine options are grayed out in vSphere Client — Broadcom Knowledge Base 344709, Updated On: 06-17-2026, accessed 20 November 2026
- Troubleshooting issues resulting from locked virtual disks — Broadcom Knowledge Base 316576, Updated On: 04-30-2026, accessed 20 November 2026
This article was originally published on .