All users disconnect simultaneously

[Mercutio879b]

We are setting up our View 5.2 system this week. We are using windows 7 x64 Enterprise. We only have about people on the system right now, but all are experiencing the same issue. Randomly all will drop their connections to their desktops. it happens 3 or 4 times a day. Not enough too be a show stopper, but enough to be seriously annoying.

a few bits of info that might be relevant:

we are using Imprivata OneSign for SSO

trend Micro AV (not deep security, but their standard client with the VDI plugin.)

the connection servers are not tunneling,

all users are currently using Teradici zero clients.

THe ZCs are on a different VLAN Than the View vms, which are on a different VLAN than the conn server.

THe affected users are all on different access switches.

THis is an automated floating pool of linked clones.

ALl are PCoIP

WE honestly don't even know where to start looking.

(Sorry for the weira capitalization, the forums don't like iPads, apparently)

[memaad]

Hi,

When the user disconnects, what message does user see ?

What is the state of desktop that you see in view administrator portal

Desktop being accessed from LAN ?

When user connect it back does user get connected.

For day, test connection with RDP or test connection from any desktop instead of teradic client

What is memory and cpu assigned to desktop

what is video memory set to in virtual desktop or number of monitor set on pool

I would prefer to do this testing with dedicated desktop so that PCOIP agent logs might reveal more information about disconnect.

You can refer this link for error code in view agent logs or view client logs

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=201210...

Regards

Mohammed

[Mercutio879b]

NO message that I am aware of. They say they just see the screen go black and they are back at the Teradici login screen.

WHen they reconnect, they are back in the desktop, right where they left off.

I Haven't been able to see it in admin yet.

WE are planning on that tomorrow.

ITs set to 4 monitors and max resolution.

WE will be looking throughout the logs tomorrow.

[mpryor]

If you're seeing them all drop at the same time, my initial suspicion is the network. Some kind of firewall monitoring and blocking connections between VLANs perhaps?

[gmtx]

Be sure to take look at the logs on the zero client as well. You should see something in there about the disconnect.

Geoff

[roshnet]

We also having same kind of issue. But its not all users simultaneously, its some users randomly. When we dig we found that this happens when the VD resources highly utilized and also IE 8 is running. We tried to regenerate this with loading CPU using load generator. Even when CPU is 100% there was no issue, but when we run IE while cpu is high VD lost session. In all related logs we didnt find any unusual error.

Hope you can also see whether this happens when any particular app is running. 

rgs

[Mercutio879b]

Very strangely, the issue resolved itself. None of the users have reported any further disconnects. I'm not sure why or how.

[Edward_Smith]

Hey all, I know this thread is old but I'm struggling a bit!

I'm having a similar issue to the above on a customer site. All users will disconnect at the same time, and again when they log back in they are right where they left off.

The view infrastructure was implemented last August, however the disconnects have started happening only since Nov/Dec. No environment changes were made around that time. There are 2 sites which are configured exactly the same, and both sites receive disconnects (albeit not at the same times).

We have opened a case with Teradici (However they have been unable to find anything conclusive on the thin client side) We found some logs stating that there was a timeout in one of the connections however.

We have had a case open with VMware who are unable to find any issues and whose only suggestion is to experiment with the NIC configurations on the ESXi servers and see if there is any effect.

We have had a case open with Netapp to check the storage, the storage is not seeing any increased IO/Network activity around the times of the disconnects. The Filers themselves are pretty idle anyway.

The only part we haven't ruled out is the network, the customer is using HP ProCurve switches.

We noticed that there is sometimes (and I mean sometimes) a latency spike on the VDI datastore, but as there is nothing going on on the Filer we have put this down to what seems like a network bottleneck, the trouble is trying to find out what could be causing it. ESXi hosts are not under a heavy load at all. There are no processes occurring around the times of the disconnects that would contribute to the issue.

Here is a short list of some disconnect times on one of the sites:

28/01/14 – Disconnect at 15:30pm.

13/02/14 – VDI disconnects at about 15:56pm

05/02/14 – Two disconnects at 14:10pm and 15:40pm.

17/02/14 – Two disconnects in same day, one at 11:10am and another at 11:48am.

19/02/14 – Disconnect at 15:15pm.

21/02/14 – Disconnect at 11:30am.

26/02/14 – Disconnects at 12:40pm, 12:45pm, 13:30pm.

27/02/14 – Disconnect at 13:40pm.

I have thought about changing some of the View settings (such as timeout values) to see if this helps but it is hard to test when the issue is so intermittent.

Can anyone think of something that may be causing any type of network problem?

Thanks in advance!

[bjohn]

http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=103069... Don't know if the above helps.