New Horizon Connection Server (8.9) Hangs and requires a daily reboot

[jlorang]

Looking for any help with this one.  Built a new horizon connection server (8.9) and all works great when it works.  Daily I have to reboot the connection server for VM's to be able to connect.  When users are not able to connect I also can not connection to the Horizon admin portal as it just times out.  After rebooting the connection server all works as it should for a day and some times two days.  Is there anything I can start to look at to troubleshoot this?

[vDruid · Accepted solution]

Hi jlorang,

I would first look at the Connection Server logs located at 

• DriveLetter:ProgramData\VMware\VDM\logs

then look at the timestamp where users are not able to connect to check if errors pops up.

Second, have you enough disk space and Virtual hardware according to VMware recommendations ?

[yukiafronia · Accepted solution]

Hi, @jlorang @vDruid 

@vDruid  is right to look at the debug log.

I would want to know when the Horizon Console becomes inaccessible, so I would keep throwing scripts at the Horizon Console all day long to check the http status every 5 minutes or so, maybe every 10 minutes.

It is normal for it to return a status of 200, but it is unlikely that throwing the script around would affect the user's connection; if the Connection Server fails just by checking the http status, then there are not enough resources in the virtual machine to begin with.

So, you should keep checking the http status towards the Horizon Console, and if you see anything other than status 200, you should look at the debug log as @vDruid  said

[jlorang]

Thank You.  The Log files is exactly what I was looking for.

[jlorang]

Hey Guys I am back as the issue continues,  I am not seeing anything jump out at me in the logs.  The CPU and Ram never show over 50% and have to reboot the server for the horizon connection portal to present.  Any advice on what I should be looking for in the Logs?

[vDruid]

Hello,

Can you share the log file to have a look ?

[Cuore_Sportivo]

Hi All,

Last month we upgraded our fully working and stable environment form v8.4.1 (2111.1) to v8.9 (2303). No errors during the upgrade, but a while after we regularly noticed that we are also impacted by the very same issue that jlorang described. Our CS becomes unresponsive -web admin console seems down while our users get a Connection Timed Out error.

The issue is resolved by restarting the VMware Connection Server service, which clearly indicates that this is a Horizon service issue.

There were no issues at all prior to upgrading to v 2303.

There is a new version (2306 or 8.10) that we will soon upgrade to, but since there is no mention of this issue in either the resolved or the known issues notes, I unfortunately don't expect much.

Nothing weird in either VMware or Windows logs, so any help is appreciated.

[jlorang]

I had a similar issue on a previous horizon environment.  Issue ended up being the default cert was left on the machine along with the friendly cert.  After removing the "non friendly" cert the view server/environment was stable.

[Cuore_Sportivo]

Thanks for the rpely jlorang.

We just created a new replica Horizon CS and brought offline the previous CS, but the issue re-appeared within just three days.

Only two certs are currently in the mmc console, the default one (with no vdm in friendly name) and the active/signed cert. It seems that we'll need to involve VMware's support in this one.

[fatsheep]

hello Sir,

I have same problem when user connection server 2303. We have only 5 VDI user. CPU 1% RAM 38% , we have create new self-cert, no issue in Certificate Management (console).

But web console admin will be unreachable after some days (4 or 6 days). 

Now we plan task schedule , restart server after 12 hours every day.

Please tell us know when upgrade to 2306, the problem will be resolved or not?

[fatsheep]

how about after you upgrade to ver 2306? @Cuore_Sportivo 

[Cuore_Sportivo]

Hi @fatsheep

We're still affected by the issue but haven't yet upgraded to 2306.

[fatsheep]

At this time, we set script auto restart service connection server after 12 hours (it mean 1 day we will restart service 2 times, I set 12h AM and 12h PM).

If this issue resolved in 2306 please let me know, many thanks !

[fatsheep]

Hello all,

We have upgraded vconnetion from 8.9 2303 to 8.11 2309 and service run stable. Don't need restart service in task scheduled.

We will monitor and feedback if have any problem in the future