Message: Failed to bootstrap VCF services runtimeRemediation Message: [Blank/None]Reference Token: #####Cause: java.lang.StringindexOutOfBoundsException: Range [-1, 0) out of bounds for length 0"Cannot complete file creation operation. Failed to create object."In the vCenter server, /var/log/vmware/vpxd/vpxd.log shows below error
vpxd[13####6] [Originator@6876 sub=vmomi.soapStub[61] opID=###-01-##-01] SOAP request returned HTTP failure; ... method: reserveName; code: 500(Internal Server Error); fault: (vim.fault.CannotCreateFile) {--> faultCause = (vmodl.MethodFault) null,--> faultMessage = (vmodl.LocalizableMessage) [--> (vmodl.LocalizableMessage) {--> key = "vob.vsanprovider.object.creation.failed",--> arg = <unset>,--> message = "Failed to create object."--> }--> ],--> file = "vcf-services-runtime-template-9.1.0.0.25370367 (Cannot Create File)"...--> Cannot complete file creation operation."
In SDDC Manager Logs, /var/log/vmware/vcf/domainmanager/domainmanager.log show below error
vCenter's ResourcePool.ImportVAppLRO -> CreateVM -> vpx.vmprov.ReserveDirectory calls ESXi host <FQDN_OF_HOST> (host-####) via vpxa's reserveName/CreateDirectory, which returns vim.fault.CannotCreateFile with fault key "vob.vsanprovider.object.creation.failed" — vSAN's object/provider layer refused to create the storage namespace object for the new VM on the vSAN datastore backing <DATASTORE_NAME> (vsan:524####a4b7721-1b9####8686##b). SDDC Manager surfaces this as [######] PUBLIC_VSP_CLUSTER_BOOTSTRAP_FAILED
VCF 9.1
VSAN ESA Cluster
The failure is caused by a structural conflict between the physical topology of the vSAN ESA Stretched Cluster and the assigned VM Storage Policy. The issue occurs when a policy is configured for Dual Site Mirroring with a local protection level of Local FTT=2 (RAID-6) on a cluster that lacks the minimum host count per site.
In a vSAN ESA Stretched Cluster, RAID-6 (Local FTT=2) requires a minimum of 6 hosts per site. In an 8-host stretched cluster (4 hosts per site), vSAN cannot satisfy these placement rules, causing the storage layer to reject the object creation.
To resolve this issue, the VM Storage Policy must be updated to align with the physical host count of the fault domains (e.g., changing Local FTT=2 to Local FTT=1/RAID-5).
For detailed, step-by-step instructions on how to modify the storage policy