All Posts

VMware Cloud Foundation 4.5 Release Notes Parallel upgrades of vCenter Server are not supported If you attempt to upgrade vCenter Server for multiple VI workload domains at the same time, the upgrade may fail while changing the permissions for the vpostgres configuration directory in the appliance. The message chown -R vpostgres:vpgmongrp /storage/archive/vpostgres appears in the PatchRunner.log file on the vCenter Server Appliance. Workaround: Each vCenter Server instance must be upgraded separately. Performing parallel prechecks on multiple workload domains that use vSphere Lifecycle Manager images may fail If you perform parallel prechecks on multiple workload domains that use vSphere Lifecycle Manager images at the same time as you are peforming parallel upgrades, the prechecks may fail. Workaround: Use the following guidance to plan your upgrades and prechecks for workload domains that use vSphere Lifecycle Manager images. For parallel upgrades, VMware Cloud Foundation supports up to five workload domains with up to five clusters each. For parallel prechecks, VMware Cloud Foundation supports up to three workload domains with up to four clusters each. Do not run parallel upgrades and prechecks at the same time.   Using the /v1/upgrades API to trigger parallel cluster upgrades across workload domains in a single API call does not upgrade the clusters in parallel When using the VMware Cloud Foundation API to upgrade multiple workload domains in parallel, including multiple resource upgrade specifications (resourceUpgradeSpec) in a single domain upgrade API (/v1/upgrades) call does not work as expected. Workaround: To get the best performance when upgrading multiple workload domains in parallel using the VMware Cloud Foundation API, do not include multiple resource upgrade specifications (resourceUpgradeSpec) in a single domain upgrade call. Instead, invoke the domain upgrade multiple times with a single resourceUpgradeSpec for each workload domain. You can also use the SDDC Manager UI to trigger multiple parallel upgrades across workload domains.

Doc :  https://blogs.vmware.com/cloud-foundation/2021/08/24/general-availability-of-vmware-cloud-foundation-4-3/ https://core.vmware.com/resource/vcf-43-proof-concept-guide   You can have multiple cluster with in a workload domains for test , prod , dev etc .  VCF Arch basically will need Minimum 4 nodes for the MGMT WLD . MGMT domain only supports VSAN . VCF supports bring your own network device.  you can always add more nodes into MGMT upto 8 nodes and for VI WLD (workload domain ) as per Vsphere maximums.  in SDDC GUI under inventory you can add additional spare physical nodes for later scaling up .  with VCF 4.3.1 you can now use your Fabric storage when you create a VI WLD as a primary storage , instead of vsan .  https://docs.vmware.com/en/VMware-Cloud-Foundation/4.2/vcf-admin/GUID-04ECE48F-5C71-4A4C-8B5A-1D3E6AA03C1C.html https://docs.vmware.com/en/VMware-Cloud-Foundation/4.3/vcf-vi-workload-domain-design/GUID-FB2D671C-BAB6-421E-A831-A7E6EC60D647.html  

1. Are  workload domains upgraded sequentially, or can they be upgraded in parallel? yes VI WLD parallel upgrade is supported.  2. I have two workload domains. The first WLD(1) has NSX-T Managers and Edge Nodes. The second WLD(2) has been configured to use the NSX-T Managers and Edge Nodes in WLD(1). Do I have to upgrade WLD(1) before I can upgrade WLD(2)? yes WLD 1 , which will upgrade your Manager and edge , and you can start WLD 2 which will upgrade nsx-t manager. 

If your vCSA/vCenter is up and running without issue. Run the following: "/opt/vmware/vcf/operationsmanager/scripts/cli/pm_lookup_passwords.sh -u administrator@vsphere.local -e PSC -n 1 -s 0" Set your vCenter administrator@vsphere.local to the password supplied. Restart vCF.

  if you can check the logs /var/log/vmware/vcf/sddc-manager-ui-app/sddcManagerServer.log   also make sure the root partition on the SDDC is full there is known issue in VCF 4.2.x where root partition gets full which will cause this login issues .  SDDC Manager uses SSO from PSC . Make sure the Services are all start and PSC and VC is up and running fine.   

Hey, hope you are doing well. I have the same issue, right now it's with VMware Engineering team. So far what i know is that the certificate keystore got emptied  You can validate by running this command: keytool -list -v -keystore /etc/vmware/vcf/commonsvcs/trusted_certificates.store ls -lh /etc/vmware/vcf/commonsvcs/trusted_certificates.store If the command returns a 0 it means the keystore is empty.  then you can try a restore from a backup or snapshot (in case you have)  Else escalate with VMware.