vpxd
logs show delays in processing tasks. For example:/var/log/vmware/vpxd.log
YYYY-MM-YYThh:mm:SS info vpxd[23891] [Originator@6876 sub=VmProv opID=5XXXX7-01] Created VM vim.VirtualMachine:vm-XXXXX (ds:///vmfs/volumes/vsan:52############4-8##########4/9####4-3##2-6##f-00ad-1#####590/test1.vmx) on host [vim.HostSystem:host-#####,<FQDN ESXI host>] (FQDN ESXI host)
YYYY-MM-YYThh:mm:SS info vpxd[23907] [Originator@6876 sub=vpxLro opID=39####7de] [VpxLRO] -- BEGIN task-124050 -- vm-XXXX-- vim.VirtualMachine.reconfigure -- 52XXXb-836c-4XXa-cXX5-5XXXXXXa(5XXX9-f3f9-1XX1-47c2-85XXXXXX25f)
VMware vCenter server 7.x
This issue is caused by problems with the HTTP/2 protocol used by Envoy to communicate with the vpxd service. While vpxd acknowledges the requests from Envoy, it fails to process them due to HTTP/2 handling issues, resulting in tasks getting stuck or delayed.
This issue is resolved in vCenter Server 7.0 Update 3i (build number 20845200) and vCenter server vCenter Server 8.0a (build number 20920323) onwards. To download the vCenter server ISO follow the link Download vCenter Server patches and ISOs for latest releases
Workaround: -
Note: Before proceeding, take a powered-off snapshot of the VCSA VM. If vCenter is deployed in Enhanced Linked Mode (ELM), take powered-off snapshots of all vCenter servers in the environment.
Log in to the vCenter Server Appliance (VCSA) via SSH.
Back up the existing configuration file:
Modify the configuration
vi /etc/vmware-rhttpproxy/endpoints.conf.d/vpxd-rhttpproxy-endpoint.conf
Change the following lines:
From:
To:
Restart the rhttpproxy service to apply the changes: