As a View administrator, you would notice multiple events in the EVENTS of the VIEW ADMINISTRATOR.
"The pending session for user <domain/user> on VM <vm name> has expired".
If you are seeing this issue on a floating pool where the VMs are set to refresh or delete on logoff, this issue can be caused if your DNS server is not getting updated with the new IP address of the newly created/refreshed VMs.
When a View VM is deleted or refreshed, it gets a new MAC address and thereby a new IP address. Now ideally the DHCP server should update the DNS server. But in some environments, this does not appear to be the case. You can verify this by checking the IP on the VM itself and in the DNS server.
If you are experiencing this issue, you need to make a change on your DHCP server. For the scope under which the VMs fall, make sure the below option is enabled.
"Always dynamically update DNS A and PTR records"
If this helps in fixing your issue, please do leave a comment.
And yes, also make sure that the userinit key on the View VM has got the correct value. You can refer to VMware KB article or communities for this detail.
If this article is of some help, please leave a comment.
And yes, also make sure that the userinit key on the View VM has got the correct value. You can refer to VMware KB article or communities for this detail.
If this article is of some help, please leave a comment.
Four years old, and still a useful post. Nice work
ReplyDeleteIndeed. This article is 4.5 years old and it just made my day. After reading this I realized that this issue started for us when we switched DNS servers. Very nice catch. Thank you.
ReplyDeletestill issue in my environment above mentioned DNS setting is already Active ... still having multiple sessions
ReplyDeleteAre you still having this issue? We are seeing this now and our DNS Setting is set correctly.
DeletePerfect.
ReplyDeleteDoes anyone know if this setting is worded identically in an Infoblox environment? I took a very brief look at our VDI subnets in there (I'm not an Infoblox admin), and I didn't see that setting. Trying to get our admins to identify it, but, would love to know if it's identified differently. Thanks in advance.
ReplyDeleteStill works. Also, I had set up 1 day leases because after rebuilds of our desktops they grab new IP's since they get different NIC's with new MAC's and I didn't want the IP pool to fill up with old entries, but when the lease expires while they are connected then they try to connect again it would fail due to caching the old lease. Still searching for a good solution to that but in the mean time I've set leases to infinite and wrote a script to ping hosts on that subnet then delete non-responsive entries.
ReplyDeleteI have a similar issue in our VMware Horizon RDS App environment. We run a farm of RDS servers, and have very intermittent behaviour where the user is able to authenticate, but launching the app causes a black screen. The horizon event logs show 'Session Expired...' for user - (this always happens exactly 1 minute after authenticating - we use Pre-launch in horizon fyi). On the RDSH server, the event log shows '"The pending session for user on VM has expired". and.or 'RPC server unavailable'.
ReplyDeleteWe're stumped, sometimes it works, sometimes it doesn't. The instances when it does work, a pre-existing pre-launch session appears to alreayd be in place on the server (despite there being zero session in Task Manager > Users - so I don't know where/how these 'Pre-launch' sessions are being created?).
Any help would save my skin!