Yes, the privileges are scoped to a certain feed. You're right, As soon as i scope to "All Feeds", login works. That's great! I'm still abit confued about how to limit some people to only one feed. They should be able to Add, View, Download packages of one feed only (including uploading packages using the web ui). Other feeds should not be visible to them. How to achieve that without scoping my privilege to a certain feed. What I didis to create a new Role "HomePageViewer" Role with only the General_ViewHomePage Task. Then, I've created two assignments for my ProductDevelopers group: Principal:ProductDevelopers, Role:Package Consumer and Adder, Scope: MyFeed Principal:ProductDevelopers, Role:HomePageViewer, Scope: (System) Would that be a recommended way or did I think too complicated?

We've finally tracked down what was causing this issue. We had two feeds configured that both had two connectors that connected back to the same ProGet server on different feeds. Turns out that this causes a very high request queue on the ProGet server, which in turn caused timeouts and eventually errors during package restore. I've now created a separate web application in IIS hosting the same ProGet folder but running on a different port and configured those two feeds to connect to the other port. This seems to alleviate the problem for now. Would be nice though if this was supported by ProGet itself. I've seen many artifact repositories that support such a feature where sort of virtualized feeds can be created by combining other feeds.

if you specify the /LogFile=[filename] argument, then the install log will be written there.

Hi, I'll try to answer the various questions here... When that publish completes successfully in TC, can we easily 'start' a build to DEV integration environment. Yes, for this , you may want to use a "URL-based Trigger" to import the build files from TC. Lastly, can we schedule a PROD release once we are done for hte middle of the night and have it alert us via text if it fails? You can schedule the promotion / deployment, and set up an event listener to email you on failure. These are extensbile, so if the SMS is needed (instead of email), you can either have a script call an SMS gateway or write an extension to do that. In addition, we don't like the non-graphical configuration, etc of bathc files as it is not overly intuitive (but works fine). Please let me know how this would work in your environment. Sure, you can run a batch/cmd file... unless it's prompting for input (like by using PAUSE command) or doing other things that can only work interactively. But, you'll find out quickly what works / doesn't. Sometimes it will be easier just to use the built-in actions in BuildMaster. Originally our organization mandated octopus, but has since undone that mandate, so I'm trying to see what options I have...The challenge is in general we have it working, so cost may be the reason we don't convert, but that is something to understand after a POC. For a comparison check out A Comparison: BuildMaster vs. Octopus Deploy .

This is likely an incorrect error that's being reported; next time you see this, just try to restart the web app by going to, Admin > All Settings > Save. This will re-establish a connection to the service on restart, and should clear this error.

This is fairly easy to reproduce -- just turn off the SQL Server. As to why it's happening, that's hard to say... perhaps it's being rebooted? Or you have network connectivity issues? If this is happening, I wouldn't disable pooling --- instead, just try to restart the web app by going to, Admin > All Settings > Save. This will re-establish a connection to the service on restart, and should clear this error. Another thing you may want to do, is to configure the service to crash on database errors; then set auto-recovery. You can do this by setting a value in your service's application configuration file, "CrashOnDbFailure" to "True", and changing the recovery settings in Windows.

The installer has an issue creating the initial database due to a (recently) undocumented change on their end, but you can fairly easily work-around the issue until the installer is adjusted for this... see Question #3924.

Would love to see a VSIX feed (didn't see a way to vote for this feature).

You recommenced ADInsight previously. It is an obsolete tool that is no longer maintained. It does not work in Windows 2008 R2 nor on machines with VMWare tools installed.

In the v5 plans (which we will ship in 4.9), you can have "output variables" from your powershell scripts. But currently, you can't communicate back to the BuildMAster execution engine via a script with anything but exit codes.

Yes, that's correct; if your doing this with a custom extension, then you can use the "StoredProcs.Variables_CreateOrUpdateVariableDefinition" method directly, and pass null in where appropriate.

We don't have a built-in mechanism to do that just yet

We don't have that as a retention policy currently, but can consider adding it in the future!

You could use the API to delete one by one of course, but If you delete the packages from the store on disk, and then manually run the feed clean-up indexer (Admin > Tasks), then it will do the same.

Hi, yes -- There are quite a few users that have BuildMaster to do just that. I'm not sure about the Oracle/SOA services, but if it's something that you can write in a quick script, then it'd be fairly easy to add as an action as part of your deployment plan.

hi, I'm not sure I understand the question? But, you can set up multiple database conenctions in your application, and pick the ones you want to use... either directly or with a variable.

In my case, I was generating the "real" packages in Release build configuration but generating the symbol packages in Debug build configuration. Once I realized the symbol server wouldn't respond correctly when the build configuration differs between the packages, I published the symbol packages in Release build configuration so the packages matched and everything started working correctly. Hope this helps someone else in this same situation!

You really shouldn't use API keys with ProGet, actually. From the docs: An API Key is used by NuGet client tools (such as nuget.exe and NuGet Package Explorer) as an authentication mechanism for publishing packages. It makes a lot of sense for the NuGet Gallery, as it's a community-run site where anyone can publish packages they own. In enterprise environments, the notion of individual package ownership doesn't quite translate, nor does the usage of API Keys. With ProGet, a principal (user or group) either has privileges to publish a package to a feed, or does not. This allows ProGet administrators to more easily control access and usage through LDAP and Groups instead of through community mechanisms like API Keys. So, if they can authenticate to the feed (using basic auth or windows integrated auth... or even an api key in the form of user:pass), and they are authorized, then the can push a package.

Thanks, Dan. It looks like this resolves the issue I was seeing as well! Now I just have to figure out why my ProGet server won't do a full index of npm without having the connection forcibly closed, but at least I'm back in business.

In order to do this, you'll need to set up a separate configuration profile. you can do this in Admin > Extensions > TFS.