Known Issues and Troubleshooting for VMware Live Sync I/O
This page lists known issues with this release.
- Encryption and deduplication are not supported.
- A source VM that has snapshots cannot be replicated, because VMware does not allow plug-ins to be attached when snapshots are present.
- Linked clone VMs cannot be replicated with Live Sync I/O.
- Ensure that VSA proxies and ESXi hosts have the correct time synchronization. If not, replication might remain in a starting state or replication might fail with the error "Failed to initialize data rehydrator on source". The BlrSvc.log might contain the following error:
ESX and local time differ too much. Please enable NTP synchronization.
- If a virtual machine is replicated to a destination site and you perform a restore operation for the VM at the destination site, the replicated VM and the restored VM have the same UUID. To change the UUID, see Changing or keeping a UUID for a moved virtual machine (1541).
- Source VMs with Integrated Drive Electronics (IDE) disks cannot be replicated.
As a workaround, see the VMware KB article Converting a virtual IDE disk to a virtual SCSI disk (1016192).
- VMs that have shared SCSI controller disks cannot be replicated with Live Sync I/O.
- If the source VM has independent disks, the backup fails and the source VM cannot be replicated.
- Replication is not supported for source VMs that have virtual RDMs with non-persistent disks.
- If you change a virtual or physical RDM with dependent disks to use independent persistent disks while a replication operation is in progress, the replication operation succeeds.
- A virtual RDM with dependent disks is replicated as a VMDK.
- Replication is not supported for VMs with disks on an NVME controller.
- Only one proxy at the source site and one proxy at the destination site can be used for replication. Multiple proxies and proxy failover are not supported.
- Linux proxies are not supported.
- All VMs configured for replication must use shared cluster storage. If any VM uses local storage, VMs cannot be migrated with vMotion and the ESXi host cannot be placed into maintenance mode to install plug-ins.
- Rollbacks to earlier plug-in versions are not supported.
- If you configure destination VMs to be backed up, do not enable Changed Block Tracking (CBT) on the subclients for the destination VMs. If CBT is enabled on destination VMs, test boot and permanent boot operations fail.
- For the source site, the BlrSvc service consumes up to 300 MB memory for each replication pair.
- For the destination site, when granular recovery is specified, the BlrSvc service consumes up to 1024 MBs.
During a test boot operation or when recovery points are being pruned, the consumption increases based on the number of recovery points being processed.
- After VM migration, replication might be disrupted for a few minutes.
- After a test boot operation, disk consolidation sometimes fails for snapshots that replication creates on destination VMs.
- Uninstall operation failures for VIBs might not be detected or logged.
- CBT API incorrectly returns an empty list of changed blocks.
- If a VM is deleted while snapshot disks are open, VMDKs might be left over.
Last modified: 3/12/2019 3:43:21 PM