Issues with a non-managed RDS Linux Pool

[sappomannoz]

Hello 👋 community.

I'm struggling with some issues with manual (non-managed) Linux VM RDS hosts.

First, I can't understand if they are supported or not. The Horizon documentation states it's possible but the directions  aren't working for me, see https://communities.vmware.com/t5/Horizon-for-Linux/Physical-Linux-Desktops-possible-or-not/m-p/2889...

I have opened a SR to get an answer to that, but I seems that first level support is struggling with my question.

Anyway, by using undocumented/unsupported parameters during the agent installation I'm able to register those VMs as RDS hosts and to use them for delivery apps or desktops but there is an issue. After some days the agent is no longer reachable, it's still there in the console but unreachable. 

To re-register it I have to de-install the agent on the Linux VM and also I need to empty the /etc/vmware directory. After those steps I'm albe to re-register the VM again. I also have opened a SR for this, but I still got no answer. 

[chenyu_vmware]

Hi,

Can I understand your problem is "with your unmanaged mode installation, some days later, your agent is no longer reachable"? That's mean, your RDS desktop and app can be used in normal at first, but some days later, it can't work.

If that's case, please first check farm/pool vm status at broker UI, to make sure it's in available status. If that's true, then need Linux agent dct log for further checking.

BTW, the unmanaged mode installation parameter is similar like "./install_viewagent.sh -A yes -M no -b <broker_ipaddress> -u <username> -p <password> -d <domain_postfix> -n <hostname_fqdn> --multiple-session"

After your re-register vm (or reinstall Linux agent), you need re-select vm at farm or create one new farm for your pool. thanks.

[sappomannoz]

Hello @chenyu_vmware ,

in the broker interface is says that the agents are unreachable. I can see those lines when this happens in the agent log.

2022-01-19T15:36:51.871Z DEBUG <pool-7-thread-3> [AbstractJmsConnectionUser] AsyncSessionReporter-Test-RDS-Linux teardown called 

2022-01-19T15:36:51.873Z DEBUG <pool-7-thread-3> [AbstractJmsConnectionUser] CmsPublishingManager-Test-RDS-Linux teardown called 

2022-01-19T15:36:51.890Z DEBUG <pool-7-thread-3> [Main] JMS threads terminated 

2022-01-19T15:36:51.890Z DEBUG <pool-7-thread-3> [Main] Waiting for executor service shutdown 

2022-01-19T15:37:01.892Z DEBUG <pool-7-thread-3> [Main] Forcing shutdown with 2 items on the queue 

2022-01-19T15:37:01.892Z DEBUG <pool-7-thread-3> [Main] Forcing shutdown with 0 items on the queue 

2022-01-19T15:37:01.892Z DEBUG <pool-7-thread-3> [Main] Agent messaging is stopping 

2022-01-19T15:37:01.892Z DEBUG <pool-7-thread-3> [JMSHandler] Interrrupted while waiting for JMS to stop 

2022-01-19T15:37:01.892Z DEBUG <pool-7-thread-3> [StandaloneAgent] Starting jmsHandler. 

2022-01-19T15:37:01.892Z DEBUG <pool-7-thread-3> [Main] State: JMS_STATE_STARTING 

2022-01-19T15:37:01.892Z DEBUG <pool-7-thread-3> [Main] Initialising agent messaging for server cn=947b7456-ead4-439e-9ca8-43599ba3eed2,ou=servers,dc=vdi,dc=vmware,dc=int as member of zone null 

2022-01-19T15:37:01.892Z DEBUG <pool-7-thread-3> [Main] Main Start: agent/947b7456-ead4-439e-9ca8-43599ba3eed2, , ON 

2022-01-19T15:37:01.892Z ERROR <pool-7-thread-3> [StandaloneAgent] Failed to start JMS 

2022-01-19T15:37:01.892Z INFO  <theMonitor Test-RDS-Linux> [Main] JMS monitor running in simulated mode, stopping 

2022-01-19T15:37:01.893Z DEBUG <pool-7-thread-3> [StandaloneAgent] Failed to start JMS 

java.lang.NullPointerException: null

at java.util.ArrayList.<init>(Unknown Source) ~[?:?]

at com.vmware.vdi.agent.messageserver.JmsManager.a(SourceFile:164) ~[agentjms.jar:?]

at com.vmware.vdi.agent.messageserver.JmsManager.connect(SourceFile:226) ~[agentjms.jar:?]

at com.vmware.vdi.agent.messageserver.Main.Start(SourceFile:1271) ~[agentjms.jar:?]

at com.vmware.vdi.agent.common.JMSHandler.startup(JMSHandler.java:310) ~[agent-common.jar:?]

at com.vmware.vdi.agent.standalone.StandaloneAgent.startJms(StandaloneAgent.java:1692) [agent-standalone.jar:?]

at com.vmware.vdi.agent.standalone.StandaloneAgent.handleReconfigure(StandaloneAgent.java:782) [agent-standalone.jar:?]

at com.vmware.vdi.agent.standalone.StandaloneAgent.processRequest(StandaloneAgent.java:946) [agent-standalone.jar:?]

at com.vmware.vdi.agent.common.JMSHandler.onMessage(JMSHandler.java:105) [agent-common.jar:?]

at com.vmware.vdi.agent.messageserver.TopicMessageManager.executeTask(SourceFile:3484) [agentjms.jar:?]

at com.vmware.vdi.agent.messageserver.AbstractJmsConnectionUser$a.run(SourceFile:272) [agentjms.jar:?]

at java.util.concurrent.Executors$RunnableAdapter.call(Unknown Source) [?:?]

at java.util.concurrent.FutureTask.run(Unknown Source) [?:?]

at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(Unknown Source) [?:?]

at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source) [?:?]

at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source) [?:?]

at java.lang.Thread.run(Unknown Source) [?:?]

2022-01-19T15:37:01.893Z DEBUG <pool-7-thread-3> [ComponentResponse] Message is RECONFIGUREDONE 

2022-01-19T15:37:01.893Z ERROR <pool-7-thread-3> [JMSHandler] JMS Error:null 

2022-01-19T15:37:01.893Z DEBUG <pool-7-thread-3> [JMSHandler] JMS Error:null 

java.lang.NullPointerException: null

at com.vmware.vdi.agent.standalone.StandaloneAgent.handleReconfigure(StandaloneAgent.java:799) ~[agent-standalone.jar:?]

at com.vmware.vdi.agent.standalone.StandaloneAgent.processRequest(StandaloneAgent.java:946) ~[agent-standalone.jar:?]

at com.vmware.vdi.agent.common.JMSHandler.onMessage(JMSHandler.java:105) [agent-common.jar:?]

at com.vmware.vdi.agent.messageserver.TopicMessageManager.executeTask(SourceFile:3484) [agentjms.jar:?]

at com.vmware.vdi.agent.messageserver.AbstractJmsConnectionUser$a.run(SourceFile:272) [agentjms.jar:?]

at java.util.concurrent.Executors$RunnableAdapter.call(Unknown Source) [?:?]

at java.util.concurrent.FutureTask.run(Unknown Source) [?:?]

at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(Unknown Source) [?:?]

at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source) [?:?]

at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source) [?:?]

at java.lang.Thread.run(Unknown Source) [?:?]

This happens for example, when I reboot all my connection servers sequentially (I waiting for the broker server to work properly prior to reboot the next)  after a windows update.

I have SR 21291374412 open since weeks but we didn't make any progress.

In my installation string I use the fqdn for the broker, but it shouldn't make any difference.

I also tried with the -k parameter, the documentation is not very clear, but it should be the KDC address.

It was impossible to use the -k parameter, I always get an error, I wonder if it's still supported.

[chenyu_vmware]

Use the fqdn is also fine.

Can I understand your RDS desktop or App never works normally after your unmanaged-mode installation?

What's status of your vm at "Registered Machines" of broker Admin UI? Or could you please send your viewagent-debug.log and /etc/vmware/viewagent-machine.cfg file to chenyu@vmware.com, I would like to take a look at log file first. thanks.

[sappomannoz]

Hi @chenyu_vmware ,

the apps and desktops are working perfectly until I have to reboot my connection servers. After those reboots I can still see the registered machines under the RDS host tab, but the agent is unreachable and the users can't login into their apps/desktops

I will send the files now.

[chenyu_vmware]

Thanks for your additional information.

Now I understood your faced problem. The jms connection is broken between broker and Linux agent after broker reboot, then you can't login RDS desktop or App.

But whether it's one problem, I will try to reproduce it in my env and ask broker team to confirm whether it is expected, will Cc you.

Even if this problem can't be fixed finally, maybe I have one alterative way for you to make your RDS desktop or app work, but it's one totally another solution.

[sappomannoz]

Thanks @chenyu_vmware , keep me posted!

[sappomannoz]

Hi @chenyu_vmware ,

as additional information I have also windows RDS hosts in my environment that are manually registered and those are working as expected when I reboot che connection servers.

[chenyu_vmware]

Hi Cumer,

Looks I can't download log file from SR link, I have received the agent logs by mail, could you please send me the broker log to my mail as well. Thanks.

[chenyu_vmware]

Please request VMWare Support Engineer to open PR to dev team, for next follow up. Note that please attach both connection server and agent log files at PR. As it need broker team's help to analysis the root cause. Thanks.