No agent server was available

Hello,

I’m running a couple of ThinLinc clusters, one old and one new. I’m not very familiar with ThinLinc so I’m just trying to keep the old one going while moving everything to the new one.

My old cluster has one machine in the subcluster that refuses to give a useable session. I compared the configs via tl-config -Ra / and the two configs are near identical except for hostnames.

The vsmagent and tlwebaccess services are running. Both servers are running ancient Debian with the same kernel versions.

Here is the /var/log/vsmagent.log with DEBUG2 set:

tl-session: tl-xinit exited with status=1
2024-05-22 09:57:40 INFO vsmagent: Got SIGTERM, signaling process to quit
2024-05-22 09:57:40 INFO vsmagent: Terminating. Have a nice day!
2024-05-22 09:57:42 INFO vsmagent: VSM Agent version 4.11.0 build 6323 started
2024-05-22 09:57:42 INFO vsmagent: My public hostname is badhost.domain.name
tl-session: tl-xinit exited with status=1

The vsmserver.log on the ThinLinc server shows:

2024-05-22 10:10:43 INFO vsmserver.session: User with uid 1193820 (XXXX) requested a new session
2024-05-22 10:10:43 DEBUG vsmserver.session: client_params: {'hwaddr': 'AC91A1968289', 'capabilities': {'sound': {}, 'printer': {}, 'smartcard': {}, 'client_lang': 'en_US', 'client_platform': 'win32', 'client_version': '4.16.0', 'client_build': '3389'}}
2024-05-22 10:10:43 DEBUG vsmserver.extproc: Starting process ['/opt/thinlinc/libexec/tl-session', '--initgroups']
2024-05-22 10:10:43 DEBUG vsmserver: List of groups: [10513, 514629, 587641, 587642, 634697, 643766, 711939, 1247762]
2024-05-22 10:10:43 DEBUG vsmserver: User XXXX requesting session on cluster 'Default'
2024-05-22 10:10:43 DEBUG vsmserver: Contacting VSM Agent badhost.domain.name to request a new session for XXXXX
2024-05-22 10:10:44 WARNING vsmserver: VSM Agent badhost.domain.name reported failure creating session for XXXXX
2024-05-22 10:10:44 WARNING vsmserver.loadinfo: Agent badhost.domain.name failed to start session, Increasing penalty points by 5 to 5
2024-05-22 10:10:44 DEBUG vsmserver: Contacting VSM Agent goodhost.domain.name to request a new session for XXXXX
2024-05-22 10:10:46 INFO vsmserver: VSM Agent goodhost.domain.name successfully created a new session for XXXXX
2024-05-22 10:10:46 DEBUG vsmserver.session: Writing active sessions to disk
2024-05-22 10:10:46 DEBUG vsmserver.session: Done writing sessions
2024-05-22 10:10:46 DEBUG vsmserver: Requesting VSM Agent goodhost.domain.name to unbind ports for XXXXX's display 1
2024-05-22 10:10:50 DEBUG vsmserver: Handling connection from ('GOODHOSTIP', 50480)
2024-05-22 10:10:51 DEBUG vsmserver: Handling connection from ('GOODHOSTIP', 50490)

Didn’t see anything I could make much sense of in /var/log/syslog or dmesg.
I’ve attempted reboots and multiple accounts. I’ve even tried rerunning tl-setup, for laughs.

Where should I look next?

Worked this out with Cendio. One of my own thinlinc session xinit.logs was filling the drive to capacity with a response loop from a EULA software agreement that I wasn’t aware was running. So, clearing the log solved my issue.

Thank you to Martin and others at Cendio!

Hey,

Happy to hear that you managed to get it sorted out, happy to assist!

Kind regards,

Martin

1 Like