Hi,
I have 2 SSH and one SCP task in my deployment plan, all connecting to a server TJiraHost.ac.local on the local network using port 22.
These tasks seem to fail 99% of the time and then occasionally they succeed.
Whats even more interesting is that whilst this is failing, if I manually SSH using putty from the box on which the Bamboo server is running to TJiraHost.ac.local on port 22 (using the same login details) the putty connection succeeds - every time.
Therefore, I know the hostname, port, username and password are correct, but I cannot figure out why Bamboo fails whilst putty succeeds.
At first I thought it may be related to the keep alive so I set this to 5s for each task and still faced the same issue.
Just for completeness, TJiraHost is running CentOS and the Bamboo server is a Windows server.
This ended up being purely a network based error.
Essentially, the builds were occurring 99% of the time on a remote agent, from which the SSH connections would also then be attempted. However, this remote agent was not able to see the deployment server due to subnet issues.
The other 1% of the time the SSH connections were made on the local agent (the machine from which I successfully tried SSH using putty). This local agent was able to see the deployment server as it had visibility of both of our subnets. This meant in these 1% of deployments the SSH tasks passed without issue.
I'm facing the similar issue. In my case I want to ssh to one of my EC2 instances on aws from Bamboo's local agent. Both are inside the same VPC. I can ssh to my EC2, if I'm logged into Bamboo's local agent. But, I want to achieve the same using ssh task in Bamboo. Do you think Imight also have an issue with the subnet's settings? How can I check that?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.