We started evaluating octopus on a dev server (TFSD1) and that went well, so we promoted it to a production box (OCT01). We followed the standard procedure of migrating to a new server, but I think I’m running into a communication problem.
The octopus server config files between the two servers share the same SQUID value. Both servers also have the same thumbprint, and because data was migrated, some of the same machines exist in both deployments (i.e. IISD1 and IISD2). TFSD1 is on a trial license while OCT01 is on a paid license.
I feel like this is the cause of some of the dropped communications I’m seeing between a tentacle on IISD2 and OCT01. Additionally, on OCT01, I’m seeing pending health checks for TFSD1 and I’m seeing “Could not re-lease subscription license” errors in the RavenDB log. I’m thinking that this is somehow related to TFSD1 being on a trial license.
Not really sure how to iron this out. Any advice is greatly appreciated.
Thanks for getting in touch! Let’s split this up a bit to start the troubleshooting:
The message "Could not re-lease subscription license"it’s an internal RavenDB licensing call which isn’t relevant because we have an embedded license, but it still tries to call out. You can safely ignore this one
The trial licence vs paid licence should not be a problem either, as long as the trial is still valid, meaning it allows you to have enough Projects, Tentacles and Users to support your backup.
Now for the Octopus server migration:
The comunication issues you are having is because both instances are sharing the same SQUID. This value should be unique for each instance, or otherwise the Tentacles could get confused and start dropping connections. So the first thing you’ll need to do is stop the old Octopus server instance (TFSD1). Once this is done, go to the new server (OCT01) and follow these steps:
Stop the Octopus Server service
Find this file and open in it a text editor: C:\Octopus\OctopusServer\OctopusServer.config
Update SQ-WIN2014R2-0D9E47ED with a unique SQUID. Changing a few characters should be enough.
Save and close the file
Restart your Octopus Server Service
On the environments page, reset Tentacle connections.
This way OCT01 will have a new SQUID and it will use it to stablish a new handshake with the Tentacles.
If this doesnt work, it might be needed to make some manual modification on the config files.
So I have done that, only instead of SQ-WIN2014R2-0D9E47ED, I chose a different SQUID that more closely matches the server name.
One thing I am noticing now is that list of orchestrated activities is getting larger. I don’t think that’s right. Plus, the steps that I have to send post deployment emails are no longer working.
Error in EmailActionOrchestrator-B+Y-AjL5aK80pA@SQ-CINSHROCT01-0D9E50FE while receiving {"Id":"e817dff9-270e-11e5-80e2-005056af2cba","From":{"SerializedValue":"DeploymentStepOrchestrator-B+U-AjL5aKpt3w@SQ-CINSHROCT01-0D9E50FE"},"Headers":{"Created":"2015-07-10T14:21:13.7822876+00:00","Expires-At":"635722212737822876"},"To":{"SerializedValue":"EmailActionOrchestrator-B+Y-AjL5aK80pA@SQ-CINSHROCT01-0D9E50FE"},"MessageType":"Octopus.Platform.Deployment.Messages.Deploy.Email.StartEmailActionCommand","Body":{"$type":"Octopus.Platform.Deployment.Messages.Deploy.Email.StartEmailActionCommand, Octopus.Platform.Deployment","Logger":{"LoggerActorId":"Octopus.Logger@SQ-CINSHROCT01-0D9E50FE","CorrelationId":"ServerTasks-7162/75f8ef7e48024290ade65391d9bcefe6"},"DeploymentId":"deployments-1239","DeploymentActionId":"f3f25da9-2ab9-40be-a39c-c235b6a25c1d"}}
System.Net.Mail.SmtpFailedRecipientException: Mailbox unavailable. The server response was: 5.7.1 Unable to relay for dev@domain.com
at System.Net.Mail.SmtpTransport.SendMail(MailAddress sender, MailAddressCollection recipients, String deliveryNotify, Boolean allowUnicode, SmtpFailedRecipientException& exception)
at System.Net.Mail.SmtpClient.Send(MailMessage message)
at Octopus.Server.Orchestration.Deploy.Email.EmailActionOrchestrator.<ReceiveAsync>d__0.MoveNext() in y:\work\refs\heads\master\source\Octopus.Server\Orchestration\Deploy\Email\EmailActionOrchestrator.cs:line 53
EDIT: Granted the e-mail problem may just be a coincidence. We’re investigating that.
The tentacles have all connected successfully and their health checks are good.
There are quite a few pending health checks for TFSD1 which is the old octopus server that we migrated from. If I go into the configuration menu, select diagnostics and then click on view details they will all be listed there.
Sorry for the confusion! You can cancel the tasks that you are seeing on OCT01 that refer to TFSD1. If OCT01’s Health Checks are running successfully, then everything should be ok.