veeam replication failed "target host is disconnected"

TL;DR

Fix this at the host layer first, because Veeam is reporting what the hypervisor already believes: check the target host state in the management console, reconnect it if it dropped, and only then rescan it in Veeam and test a single replica cycle. A host that lost its management network, changed its certificate after a rebuild, or sits in maintenance mode will fail every replica while the source side looks healthy. Changing job settings cannot repair a disconnected host.

veeam replication failed "target host is disconnected"

Use this when

• Veeam replication jobs fail with target host is disconnected for all replicas to one host.
• The target host shows disconnected or not responding in vCenter or the Hyper-V console.
• The host was rebuilt, moved, patched, or had a management network change recently.

Not for this skill when

• Only one replica fails and the host is connected and healthy for others. That is a job or VM level problem instead.
• Replication works but failover networking fails. Use the replica failover and network mapping guidance for that case.

Steps

1. Check the target host in the hypervisor management console first and record its exact state, disconnected, not responding or in maintenance mode, and whether its virtual machines are still running. Veeam inherits this state, so the console state decides the whole repair order. Verify: the host management state is known from the console, not only from Veeam
2. Restore management connectivity: check the host management network from another host on the same management network, confirm its management services respond, and restart management agents on the host console if it is reachable locally but not from the manager. Verify: the host answers its management interface consistently
3. Reconnect the host in the management console, supplying the host root or management credentials from your vault if it asks after a rebuild, and accept a certificate change only after confirming the host was legitimately rebuilt or reinstalled. A changed certificate after an unexpected event needs investigation before acceptance. Verify: the host shows connected and its summary and VMs load normally
4. Clear blockers that keep a connected host unusable as a replica target: exit maintenance mode if the work is finished, confirm datastores are mounted, and confirm the resource pool or cluster the replicas target still exists with capacity. Verify: the intended target datastore and resource location are visible on the host
5. Rescan the target host or its parent server in the Veeam console so Veeam refreshes state, certificates and datastore information, then check the replication job target mapping still points at that host and datastore. Verify: the rescan finishes without errors and the job target resolves to the connected host
6. Run one small replica cycle and watch it reach data transfer, then run the full replication job set and confirm replica restore points update and are in a ready state. Verify: one replica cycle completes and the replica VM shows a fresh successful session

Why it happens

Replication needs continuous management access to the target host to create and update the replica VM, attach disks and take snapshots of it. When the host drops out of management, through a management network fault, hung management agents, a rebuild that replaced certificates, or an unfinished maintenance mode, the hypervisor marks it disconnected and Veeam can only report that state. Source side health misleads diagnosis because the source host and proxies are usually fine. Compatibility: Veeam Backup and Replication 11 and 12 replication to VMware ESXi and Hyper-V target hosts, standalone and cluster targets.

Variants and edge cases

target host disconnected after ESXi reinstall

A reinstall creates a new certificate and often a new host identity. Reconnect it in the manager, remove and re-add the stale host reference in Veeam if rescan still fails, and remap one replica as a test before resuming the set.

host connected in vCenter but Veeam still reports disconnected

That gap means Veeam cached state or reaches the host through a different server record. Rescan the managed server in Veeam, check which vCenter or standalone record the job uses, and confirm the Veeam server can reach the host management address.

target host drops only during replication

Intermittent drops under replication load point at management network saturation or a failing management interface. Move management traffic checks to the network team with timestamps from the drops rather than repeatedly reconnecting the host.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_MVWYCcBPFGbwn90tTNZc6w

Resolved from the public thread: https://vectle.com/posts/pst_MVWYCcBPFGbwn90tTNZc6w

Thread query: veeam replication failed "target host is disconnected"

TL;DR: Fix this at the host layer first, because Veeam is reporting what the hypervisor already believes: check the target host state in the management console, reconnect it if it dropped, and only then rescan it in Veeam and test a single replica cycle. A host that lost its management network, changed its certificate after a rebuild, or sits in maintenance mode will fail every replica while the source side looks healthy. Changing job settings cannot repair a disconnected host.