Portal User asked a question.
Since a while almost all of my devices are not transferring their data anymore. All checks so far have not shown any root cause
Environment: Snow License Manager 9.15, Snow Inventory 6.7
The Snow Inventory Agent Log on the client shows:
2021-09-15T18:35:54+02:00; Verbose; snowagent; win32\winhttpmessagehandler.cpp; `anonymous-namespace'::WinHttpMessageHandler::Send; Contacting server 'xxx:11000/api/client/hello'.
2021-09-15T18:35:54+02:00; WARNING; snowagent; http\configmessagehandler.cpp; Client::ConfigMessageHandler::SendToEndpoint; A http send error occured with endpoint xxx:11000. Cannot parse HTTP request url: The URL does not use a recognized protocol
The url xxx:11000 shows the standard Snow Inventory Page.
I've seen something similar where a proxy server or some form of load balancer is intercepting the traffic from the agents to the inventory server.
Do you know if something like that is in place?
No need to mention that port 11000 is kind of unusual. 😉
I have seen port 11000 mentioned in the context of Azure VM machines.
Even if the Snow Inventory website opens, there is still a difference between web browser and Snow Agent.
As Joe already mentioned, I have also seen this with a Netscaler between a Service Gateway and the Inventory server.
In addition: what is the difference between the devices transferring data and those that don't transfer?
Loadbalancer / FW or similar is not existing for users that are in the office on our LAN.
As for the port 11000, it was working for more than 1 year for 1000 devices. But something suddenly must have happened for this not to work anymore.
The only other change that might have happened around that the time was that we enforce TLS 1.2.
Need some more time to compare a device that transmits and devices that dont transmit the data.
If you can 100% say a proxy / load balancer is not intercepting the traffic and i have seen companies say this but then network teams have trace routed the traffic and found they did then yes focus on TLS changes as if you blocked other TLS versions on the server did you also make the changes on your endpoints.
i.e. old .net installs will try to use TLS 1.0 first and then fail to try anything else as they get the "this isn't working" response back from the server.
The .net version on the machines that are functioning is 'older' (4.7).
the .net verison on the machines that are not functioing is newer (4.8)
Was using
(Get-ItemProperty "HKLM:\SOFTWARE\Microsoft\NET Framework Setup\NDP\v4\Full").Release
To get the .net version.
Looks like all devices with Agent version 6.3 are functioning correctly, those that have been upgraded to 6.7 are not reporting anymore.