Use AWS Spot instances for build agents?
Our infra setup is in AWS but we use Azure DevOps for repo, boards & pipelines. To run automated tests like cucumber, we need private agent. if there's a way to utilize spot instances for builds instead of having to have agents always on?
or
can we have a Master slave setup, like how Jenkins has?
In the self-hosted agent configuration, we can set the agent to run as service, If you configured the agent to run as a service, it starts automatically. You can view and control the agent running status from the services snap-in.
If we configured the agent to run interactively, open org settings->Agent pools->select agent pool and click the tab Agents, we can see the agent status is Offline if we do not open the run.cmd, if we need to use the agent, we can open the agent file run.cmd to run the agent.
Please refer this link for more details: Self-hosted Windows agents
Related
We have CD pipelines set up in Azure to deploy to App Services and all works well but we want to add a stage to automate out Cypress test process. The problem we have is our test environment is only accessible via VPN which is fine from local machines as we run the VPN client.
Does anyone know how to include a command within the Yaml pipeline to establish a VPN connection from the pipeline host which would allow our Cypress tests to run? I'm assuming this would require a command line connection script.
We are using a Pritunl VPN server which accepts OpenVPN connections.
Thanks.
This opens up a conversation around storing the secret and the infrastructure you have to allow that VPN client in. Azure pipelines can run arbitrary commands, but you'll need to inject the VPN secrets/key, which without strong security oversight you can have some major issues down the line.
I'd take a step back and revisit your options here, maybe build the test/ environment in azure, so you don't have to worry about this?
I am very new to consul , and has been reading about consul clustering recently. My understanding is , for each node (equivalent to a physical machine or VM), we will run a local consul agent (in client mode), hence any microservices running in that node will register itself thru this agent. but what happen if this one and only one agent is down, won't the microservices in that node unable to register anymore? Or should we expect more than one consul agent (in client mode) per node to handle such situation?
You are correct. If the Consul agent is down, the services on that host will not be able to register with the agent, and Consul will consider all services which were previously registered against the agent to be unavailable.
A very simple solution is to run Consul under a process manager like systemd, and configure systemd to restart the agent if the process unexpectedly fails. You can find an example systemd unit for this at https://learn.hashicorp.com/tutorials/consul/deployment-guide#configure-systemd. If Consul is installed from the HashiCorp Linux package repo (https://learn.hashicorp.com/tutorials/consul/get-started-install), this systemd unit will be included as part of the installation package.
I don't understand but how do I set in the hosted azure pipeline build server to be a bigger machine with more ram and more CPU's?
i want to avoid installing a self-hosted agent on one of the Azure VM's I just want to use more stronger hosted agent, where can i configure this?
I'm using a hosted MAC agent.
This is not possible. Please check this documentation
If Microsoft-hosted agents don't meet your needs, then you can deploy your own self-hosted agents or use scale set agents.
We want to collect metrics from machines running AWS lambda in AWS. How can I get access to these machines and get DD agent installed on them.
Lambda is serverless. Datadog agent is for the host. While running lambda you have absolutely no control over the host as you are not managing it.
Hence, You can monitor application running on lambda using datadog integration of lambda for the different application.
You may follow below link for AWS Integration of datadog.
Ref: https://docs.datadoghq.com/integrations/amazon_lambda/
I have set up a Server and a BuildAgent on Amazon Ec2. I have configured the build agent to sit in the /home/ec2-user/build-agent directory (will this cause problems due to permissions?)
I then configured the build agent to point to the server through the public dns and port 8111. I have then started the build agent.
I used the Cloud Tab on the TC Web interface to setup the Build Agent on the server. I can see on the build agent host machine that a connection with the server has been established. Yet in the web interface I can see no build agents in either the unauthorized tab or the compatible tabs.
In short I cannot connect a build agent to the server?
Any help or ideas on this one would be greatly appreciated.
many thanks
Dermot