I have installed quartz.net as windows service and configured it with a job and trigger in quartz_jobs.xml. This requires the job dll to be present locally under quartz server. If my job is on a remote server, how do we specify that in the quartz_jobs.xml? Do I still need to copy the dll to quartz server?
The dll with the job must be present on the Quartz.Net server and it must be accessible by the server as well. Usually you want to copy it to the same folder where the quartz dll is.
Related
My goal is to automatically deploy EAR file in WebSphere Application server cluster by Monitored Directory Deployment. So my deployment target is cluster. WAS version is 9.0.0.10. Everything works fine if I drop EAR file in
monitored directory for example '/{monitored_directory_defined_in_WAS}/clusters/my_cluster_name/'. Application is deployed and started. But I also want to deploy that application into the IBM HTTP server(resides in the same WebSphere cell with cluster) in the same automatically process by Monitored Directory Deployment.
I tried to manually predefine deploymentTargets(cluster and ibm http server) in deployment.xml file, put it in the EAR file, and drop EAR in the '/{monitored_directory_defined_in_WAS}/clusters/my_cluster_name/', but WAS deploy EAR only in cluster. As a consequence I must manually via WAS console map all modules from EAR to ibm http server, but I do not want to do it manually.
My second idea/attempt was sto create separate monitored directory for the ibm http server '/{monitored_directory_defined_in_WAS}/servers/my_ibm_http_server_name/'.
First I drop EAR into the '/{monitored_directory_defined_in_WAS}/clusters/my_cluster_name/' and right after that I drop EAR in to the '/{monitored_directory_defined_in_WAS}/servers/my_ibm_http_server_name/'. Result is that the EAR modules are deployed only in to the web server, and that is not my goal.
Is that even possible by WAS Monitored Deployment Directory functionality?
Is it allowed to manually create the deployment.xml file and add it to the EAR file?
First of all installing via monitored director is not recommended in the production environments, as it lacks control.
As you correctly suspected it is not possible to install it to the cluster and web server - check Installing enterprise application files by adding them to a monitored directory.
Because you can use only one server directory, drag and drop to map
applications to combinations of servers is limited. Scenarios
requiring use of more than one server, such as mapping to an
application server and a web server, are not supported by direct drag
and drop of an application file.
However, you still want to use it, you may deploy property file into monitored directory. That property file can fully customize your deployment e.g. configuring also module to map to web server.
Check Installing enterprise application files by adding properties files to a monitored directory for more details
UPDATE
If you have issues, I'd suggest the following approach - install your application 'classically' via admin console and map it to both web server and cluster. Then run wsadmin command to extract propertes:
AdminTask.extractConfigProperties('[-propertiesFileName myApp.props -configData Deployment=MyApplication -options [[SimpleOutputFormat true]]]')
Try to use format from that exported file for your properties
I had time to run it in my environment. I have app with 2 modules inside, one module is just mapped to the cluster, other is mapped to both cluster and web server. Here is relevant part of the property file:
taskName=MapModulesToServers
row1={ module=HelloTestUI #readonly
uri=HelloTestUI.war,WEB-INF/web.xml #readonly
server=WebSphere:cell=!{cellName},cluster=!{clusterName} }
row0={ module=HelloTestWeb #readonly
uri=HelloTestWeb.war,WEB-INF/web.xml #readonly
server=WebSphere:cell=!{cellName},cluster=!{clusterName}+WebSphere:cell=!{cellName},node=!{nodeName},server=!{serverName} }
I didnt try to use that property file to deploy the app via monitored directory, but as you can see the entry is created and mapping is done via + sign that connects cluster and web server.
If you dont see the mapping to your web server, make sure you saved the changes done in the console, and then connected via wsadmin, otherwise wsadmin will have not current data.
I'm running an ASP.NET Core web service that's dead easy to deploy to IIS on a Windows Server using WebDeploy: Right-click the project in VS, choose Publish, done. The currently running site is backuped and then updated. Couldn't be easier.
However, Windows services are another matter. I'm taking over maintenance of a service that is currently updated by building the Release config, remote desktoping to the server, stopping the current service, renaming the folder the service executable resides in (for easy rollback in case something goes wrong), copying over the build output from the dev machine, renaming the copied folder to the service folder name so the service finds the executable, and starting the service.
Is there a way to publish/update Windows services that is as easy as running WebDeploy? I haven't found any. There's always the possibility of abusing WebDeploy and just replacing the Windows service head project with an ASP.NET head project running on IIS, but the service has nothing to do with the web and I have no idea whether there would be any side effects to this solution.
(I could of course just write a script that performs the copying etc., but a built-in VS deployment process as with web projects would be optimal. As for continuous integration/deployment, it is unfortunately not something we have ever done at my company.)
I have ASP.NET Core web application that is hosted in IIS. We also have build server used for continuous integration using Jenkins. Jenkins get latest from GitHub and builds & publish the project in local folder and then execute powershell script to copy published folder to web server. ( it simply copies the files from local folder to network folder using Copy-Item)
The powershell script first stops the remote IIS web site before copying the file. (and i have confirmed that the remote web site is indeed stopped before deploying)
However every now and then we get error while copying files
Copy-Item : The process cannot access the file
\\some_ip\_apps\myapp\UI\UI.dll because it is being used by another
process.
So i have downloaded handle to check who is accessing this file on web server and i turned out its dotnet.exe
How do i stop dotnet.exe remotely or how do i release the handle remotely using powershell
I agree with the other comments about using WebDeploy to simplify your approach a little.
As far as powershell goes the Get-Process and Stop-Process along with normal Remoting stuff will let you shut down dotnet.exe.
I created a sample project using VS2017 "ASP.NET Core Web Application" with Angular.
Then in VSTS I built using Build Template: ASP.NET Core (Build ASP.NET Core Web Application targeting .NET Core). This built successfully and in the artifact section kept the build in a .zip file.
I would like to deploy it to a scale set. So for the Release section I used "Azure Virtual Machine Scale Set Deployment" template. While filling it out there is a field "Deployment Script field" which is mandatory. Not sure what this script is used for and how do i configure in it and where to source the script from.
TL;DR:
This script deploys the content of your deployment package on the virtual machine that you're preparing. In the case of an Angular ASP.NET Core app, this means that you will have to install IIS, the .NET Core Windows Server Hosting Bundle and copy your app to IIS.
Extra info
The Azure Virtual Machine Scale Set Deployment template creates an environment with two tasks:
Build immutable machine image
Azure VMSS: update with immutable machine image
Both tasks have a bunch of fields that need to be configured. Release Management allows you to link important fields to your process. This means that these settings are all grouped under the environment and can be easily set without going through the tasks.
If you look at the following screenshot you'll see that beneath the selected Deployment script field there is a line: This field is linked to 1 setting in 'Build immutable machine image'.
If you look at the documentation for the Build immutable machine image tasks you see the following for the Deployment script:
Deployment Script: Specify the relative path to the PowerShell script (for Windows) or shell script (for Linux) that deploys the package. This script should be within the deployment package path selected above. For example, Deploy/ubuntu/deployNodejsApp.sh. The script may need to install Curl, Node.js, NGINX, and PM2; copy the application; and then configure NGINX and PM2 to run the application.
So this is the script that will automatically deploy your application to a VM. In the case of an Angular ASP.NET Core app, this means that you will have to install IIS, install the NET Core Windows Server Hosting bundle and copy your app to IIS.
I would encourage you to first create an empty Windows Server 2016 VM and experiment with creating a script that prepares the VM and installs your app. Once that's done, you can reuse this script in the Build Image task in VSTS to automate VM and image creation and use that image for your Scale Set.
UPDATE
Your VSTS Build produces an artifact. In your scenario, you are already publishing the zip file with your web application. You can configure the Publish Build Artifacts task to also publish any deployment scripts you want to use during your release. These scripts should be stored in version control. The whole artifact, including your zip file and any deployment scripts you add, is used as input to your Release Definition. The zip file contains your web app as a Web Deploy package that you need to deploy to IIS. This piece of documentation explains how to install Web Deploy and use it to deploy your app.
All the deployments in my work place are manual. I have been looking at ways to automate all of this using MSBUILD and MS Web Deploy. What I have managed to get working is as follows:
build/compile solutions
run database migrations
deploy directly to website using ms web deploy
All from one MSBUILD script.
This differs slightly from the manual process, as when a new version of a website is deployed manually, it is put into a new folder that is datetime stamped. And then IIS is pointed to the new folder.
My question is, how would I do these last 2 actions? i.e. write to a specific folder that sits a level higher up than the folder the current website points to, and then repoint to the new folder IIS.
Web Deploy does not support this functionality directly. Thankfully it does provide the runCommand provider, which is what you'll end up using.
%windir%\system32\inetsrv\appcmd.exe is the utility you'll use to swap site directories. An example of usage can be found here: How do I change the physical path of web site in IIS7 with APPCMD?
Create a script that calls appcmd. This script will always sit on your destination server. You can either decide to include it as a part of your site deployment, or move it along into the new directory from the old one. The "web deploy user" on the destination machine needs to have the appropriate credentials and access to create directories.
In your MSBUILD script, prior to deployment, call the script on the destination server. E.g.
msdeploy.exe -verb:sync -source:runcommand="C:\path\to\wwwroot\bin\script.bat"
-dest:auto,wmsvc=https://contoso.com:8172/msdeploy.axd,username=%username%,password=%password%
Then deploy as usual
NOTE:
Try to avoid this way of deploying entirely. If you're looking to snapshot your site, consider deploying to a package (zip), and then deploying that package to your destination server. Rollbacks should be handled by deploying an older package, not by repointing to an older directory.