Recently we started getting some weird problems on some of our machines with octopus not finding the servers properly etc.
So I decided to re-install all tentacles on those servers.
Now I noticed something new though, under Configuration in the Diagnostics view we get messages spamming this:
Rejecting connection: the client at myip:19542 provided a certificate with thumbprint 4XXXXXXXXXXXXXXXXX14331700F6011BC827EEEA, which is associated with MyServerName (machines-580), but not configured for distribution
Rejecting connection: the client at myip:19543 provided a certificate with thumbprint 4XXXXXXXXXXXXXXXXX14331700F6011BC827EEEA, which is associated with MyServerName (machines-580), but not configured for distribution
This is a polling tentacle. and all calls seems to have a different port to it.
Anyone got any idea why this is happening, and also have a possible solution for it?
Thanks for getting in touch! To start with, can you restart the Octopus Server, and see if this resolves.
If it doesn’t a couple of questions and requests for information.
Are these listening or polling tentacles? Are they the same machine with multiple instances?
Can we get a screenshot of your environments page.
Should both of those IP addresses and ports be associated with the same machine?
Tried restarting the octopus server, nothing different, we have about 50 other polling tentacles and about 100 listening that are working properly.
This is a polling tentacle (It’s only 1) with one instance.
The IP address is associated with the machine, I’ve never seen the ports before (There is a long list of ports ranging from 19000+ that are being used, looks like they are increasing by 1 for every retry it makes.
Attaching an image of the tentacle from the octopus server.
When doing a health check we get this message:
Machine XXXXXX uses the same endpoint (physical Tentacle with SQUID ) as other machines.
We uninstalled all tentacles and the octopus tentacle manager from the server, reinstalled, even ran the new-squid command from the Tentacle.exe.
These machines were once in the environment fully working (Then removed and re-added).
Have a look to see if you have a deletion task for that machine that hasn’t properly run or timed out.
When we delete we mark it as deleted in the database but also run a task to remove the document, and sometimes that task times out.
I wonder if you have two machines for that one machine, and that is the issue here.
Hi,
Am I supposed to find this in in the tasks view - /app#/tasks - Or on the tentacle server?
I can’t find anything in the tasks view.
In the environment view there is only 1 machine for that machine.
Is there any way to purge or re-run the deletion from the server to make sure it’s gone?
To clarify, that message about multiple endpoints is from the Octopus Server side, not the Tentacle side.
We can try and find it in Raven. Under the Octopus Manager is “Browse RavenDB”.
Documents -> Machines is where the tentacles are listed.
You will need to find all machines with the same SQUID.
You can find a machine’s SQUID on it’s connectivity page.
In the machines list I can’t see any SQUID. Though I can find the machine with status “Offline” and IsDisabled = false.
The only thing differing the non-working machine with a working one is the “Version” under the Health column which is “null”
AHH that might be why it’s shared, cus its null!
So okay lets finish this off. Delete the machine from Octopus.
Log on to the tentacle and we are going to register the polling machine via command line
Use the following command. You are going to have to replace nearly everything here that is in capitals but also the admin username if its not administrator
I replaced everything and ran the command. It ended up looking like this:
“C:\Program Files\Octopus Deploy\Tentacle\Tentacle.exe” register-with --instance “Tentacle” --server "http://MyServer.se:8081 --environment “Production” --name “MyName123” --username “MyDeployUser” --password “MyDeployUserPassword” --comms-style “TentacleActive” --server-comms-port “10943” --role “my-role”
Replaced sensitive information, but it looks something like that.
I got this error:
-------------------------------------------------------------------------------
Error: Please specify a username and password, or an Octopus API key. You can ge
t an API key from the Octopus web portal. E.g., --apiKey=ABC1234
-------------------------------------------------------------------------------
