At 16:16 I said that I created separate account for the agent service, so this is something you need to do beforehand (sorry, I didn't show that during this video), so you'll know the account name and password you provide. In my case the (local) account is "agent".
Very helpful tutorial, thank you. I wonder if you could answer a question I have about self hosted agents...... If I want to deploy code to several servers in my environment is it possible to have just one self hosted agent in my environment that deploys the code to my target servers? or do I need to have a self hosted agent on each of my target servers? The reason being I need to limit external connections from my environment to a minimum. So if I could have an 'intermediate' server that connects to azure devops with the self hosted agent and deploys to other servers that don't need a direct connection to azure devops that would be my ideal solution. Many thanks
It is possible to have just one self-hosted agent in your environment that deploys the code to multiple target servers. You can configure the agent to run on a dedicated machine or a virtual machine and use it to deploy the code to the target servers using appropriate deployment tools and scripts. However, scenario that may require the use of separate agents per environment is when the target servers in each environment have different configurations or requirements. For example, if you have a development, staging, and production environment, the servers in each environment may have different hardware or software configurations, security requirements, or network settings.
It depends what would be the goal of your agent. In my case (demo purposes only), I gave Local Admin Privileges to the account ('agent'), plus "Log on as a service" permission for obvious reason. The agent should have access to all necessary folders (for example: agent working directory), so you should be more restrictive in your enterprise set up remembering the agent will execute code it downloads from external sources. More details: learn.microsoft.com/en-us/azure/devops/pipelines/agents/windows-agent
Dear Kamil, I have basic understanding of Azure DevOps. I would like to know how the pricing is calculated for Azure Pipeline. I have 18 pipelines to be created. In Azure pricing calculator, I could see the pricing is for Number of Jobs/Pipelines. If I have 18 pipelines with each has 1 job, Only 1 pipelines might have upto 4 parallel run jobs. Do you know how much does it cost? Or guide me how to calculate. Then I can do it
Check out the prices here: azure.microsoft.com/en-gb/pricing/details/devops/azure-devops-services/ What I can see - there is extra $40 per parallel job (with unlimited minutes), so in this case it would be $160/month.
@@KamilNowinski , thanks. Yeah, earlier I thought there will be charge for every pipeline, but after few testing I figured out. Only parallel job costs. Not pipeline.
Have you ever noticed dude never shows a pipeline building a project? The real work is installing tools such as runtime & sdks on the agent that no one shows including this video
Nazik, I'm not sure what do you mean? Do you want to see how to install some packages or libraries into the agent? Give me an exact example, but I think it's not a real challenge as you just need to do it as if you were do it on local machine.
Thanks Kamil, just what I needed!
I'm glad it helped, Barney!
I would have never figured out to switch from ISE to PS prompt on my own. Thanks :)
Great to hear!
This was really great, thanks for the help.
Glad it helped! Thx.
Great video Kamil!! What password did you use for the '.\agent' account on the timestamp 16:53? Is it the same PAT that was created earlier?
No, it's not PAT from Azure DevOps. That was the password for the newly created local account '.\agent' to precisely control permission.
In the agent machine, it is not clear what you did after you copied the token from Azure Devops. What password was that you used there?
At 16:16 I said that I created separate account for the agent service, so this is something you need to do beforehand (sorry, I didn't show that during this video), so you'll know the account name and password you provide. In my case the (local) account is "agent".
I'm getting a Powershell error. It says it cannot find type [System.IO.Compression.ZipFile] when the UseDotNet@2 task is executed.
Raise the issue here: github.com/microsoft/azure-pipelines-tasks/issues
Very helpful tutorial, thank you. I wonder if you could answer a question I have about self hosted agents......
If I want to deploy code to several servers in my environment is it possible to have just one self hosted agent in my environment that deploys the code to my target servers? or do I need to have a self hosted agent on each of my target servers?
The reason being I need to limit external connections from my environment to a minimum. So if I could have an 'intermediate' server that connects to azure devops with the self hosted agent and deploys to other servers that don't need a direct connection to azure devops that would be my ideal solution.
Many thanks
It is possible to have just one self-hosted agent in your environment that deploys the code to multiple target servers. You can configure the agent to run on a dedicated machine or a virtual machine and use it to deploy the code to the target servers using appropriate deployment tools and scripts. However, scenario that may require the use of separate agents per environment is when the target servers in each environment have different configurations or requirements. For example, if you have a development, staging, and production environment, the servers in each environment may have different hardware or software configurations, security requirements, or network settings.
good one, can you make some videos on Static analysis in pipeline
Mahender, could you elaborate a bit?
@@KamilNowinski i mean regarding deployment of tools like sonar cube
Great video Kami, thank you! Can you also use this for creating a dedicated build machines for Azure DevOps Server?
I believe the architecture is very similar, so yes very likely.
Hello Kamil. What permission did you grant for local user agent? Thanks
It depends what would be the goal of your agent. In my case (demo purposes only), I gave Local Admin Privileges to the account ('agent'), plus "Log on as a service" permission for obvious reason.
The agent should have access to all necessary folders (for example: agent working directory), so you should be more restrictive in your enterprise set up remembering the agent will execute code it downloads from external sources.
More details: learn.microsoft.com/en-us/azure/devops/pipelines/agents/windows-agent
thank you for the video. how can i ugprade the agent version
In order to update self-hosted agents:
1) Navigate to Project settings, Agent pools.
2) Select your agent pool and choose Update all agents.
Dear Kamil, I have basic understanding of Azure DevOps. I would like to know how the pricing is calculated for Azure Pipeline. I have 18 pipelines to be created. In Azure pricing calculator, I could see the pricing is for Number of Jobs/Pipelines. If I have 18 pipelines with each has 1 job, Only 1 pipelines might have upto 4 parallel run jobs. Do you know how much does it cost? Or guide me how to calculate. Then I can do it
Check out the prices here: azure.microsoft.com/en-gb/pricing/details/devops/azure-devops-services/
What I can see - there is extra $40 per parallel job (with unlimited minutes), so in this case it would be $160/month.
@@KamilNowinski , thanks. Yeah, earlier I thought there will be charge for every pipeline, but after few testing I figured out. Only parallel job costs. Not pipeline.
Have you ever noticed dude never shows a pipeline building a project? The real work is installing tools such as runtime & sdks on the agent that no one shows including this video
Nazik, I'm not sure what do you mean? Do you want to see how to install some packages or libraries into the agent? Give me an exact example, but I think it's not a real challenge as you just need to do it as if you were do it on local machine.