UC4 experiences?

[Thomas Falkenberg]

Hi,

we're going to start evaluating deployment automation solutions for our software (techstack: mostly java, apache, tomcat, weblogic, oracle db).

Apart from the obvious puppet, chef, cfengine the solution "uc4" came up. Does anyone have any experience with it?
I'm wondering why I have never heard of it around the community (and devops days); maybe because it's a commercial tool...

If you have used or tried it out, what are your pros and cons and how would you compare it to puppet/chef?

Thanks & regards,
Thomas

[Alex Honor]

Hi Thomas,

If you are considering UC4, you might also consider Rundeck. Rundeck is an open source workload automation/job scheduler tool. Rundeck can integrate with Puppet or Chef in a couple of ways. This can be a nice combination since it divides the concerns between directed workflow automation and convergence/configuration management.

Disclaimer: I'm one of the Rundeck developers :)

Thanks

[Daniel Kushner]

Other commercial offerings:

BMC

Nolio (I'm from there)

Xebialabs

[Will Thames]

Other open source alternatives Ansible
Salt

I have no experience of salt but Ansible is really easy to get up and running.

Disclaimer: occasional contributor to Ansible

Will

[Thomas Falkenberg]

Thank you all for your hints so far. We will take a look at all of these.

Does anyone happen to have experience managing and deploying weblogic applications with one of these tools?

2013/1/10 Will Thames <willd...@gmail.com>

[Ernest Mueller]

It seems like that as a community we should be able to get to the
place where we can help people more than throwing the ever-expanding
list of tools at them. Just a thought.

Ernest

At 03:07 AM 1/10/2013, Thomas Falkenberg wrote:
>Thank you all for your hints so far. We will take a look at all of these.
>Does anyone happen to have experience managing and deploying
>weblogic applications with one of these tools?
>
>
>

>2013/1/10 Will Thames <<mailto:willd...@gmail.com>willd...@gmail.com>

>
>Other open source alternatives Ansible
>Salt
>
>I have no experience of salt but Ansible is really easy to get up and running.
>
>Disclaimer: occasional contributor to Ansible
>
>Will
>On Jan 10, 2013 4:03 AM, "Thomas Falkenberg"

[Daniel Kushner]

Hi Thomas,

Nolio (I work there) can manage deploying weblogic apps very well. We can certainly get you in contact with customers to hear first hand.

To be fair, I am sure that all the commercial vendors I mentioned before also handle weblogic.

Best,

Daniel

--

Daniel Kushner VP Marketing | Nolio C: +972 (54) 300-3461 O: +972 (9) 765-3292 #303 Find out more about Nolio: Website | Twitter | YouTube | LinkedIn

[Justin Ellison]

On Thu, Jan 10, 2013 at 3:07 AM, Thomas Falkenberg <falk...@gmail.com> wrote:

Thank you all for your hints so far. We will take a look at all of these.

Does anyone happen to have experience managing and deploying weblogic applications with one of these tools?

We have a almost a carbon-copy of the application stack that you describe, and we manage it all with Puppet.  However, you probably don't want to do your actual ear-based deployments with Puppet, or any other configuration management tool.  Application deployment is usually something you want to handle with something more suited to deployments.  The one exception is if you're building RPM's or debs out of your actual applications.  While that's considered a best-practice by many, I've found it near-impossible to do with weblogic applications.  We use Jenkins to manage our deployments.

As far as what we do use Puppet for, it installs Weblogic via an RPM created with FPM, and then uses a series of exec resources to execute custom jython scripts that setup the domain, the machines, servers, clusters, and services (JDBC pools, etc.)   It also sets up nodemanager and subscribes it to the admin server.

I'd love to share code, but I'm afraid I don't have permission.

Justin

[Ron Gidron]

I concur with @Ernest below. I hope I can help make some order here from my experience.

Puppet, Chef and CFengine are really configuration management solutions. In a nutshell the philosophy behind them is that of "I need" and what they do best is assure your environment is always configured properly no matter how many servers you have in it. I say "I need" philosophy in the sense that the agent that will 

run on your server will have a list of "requirements" that it will verify the server agains and then communicate back to the network to say "I need ..." and properly configure the server to spec. These tools are great at assuring configuration and provide different levels of "self healing" functionality. The way you work with them is by writing requirements that are "needed" for your different server configurations. (there are other pieces to these solutions but that's the nutshell). 

The second group are application release/deployment automation solutions. UC4 is in that group together with Nolio and some of the other commercial tools that were mentioned below. These are workflow centric and follow a guidebook/user manual philosophy.you this is used to automate the application deployment process, generally by designing a generic workflow for your application that is then executed onto any environment (dev, test, qa, production etc..) and deploys the application automatically. ARA solutions facilitate DevOps efforts specifically when promoting applications (either full versions or smaller sprints etc..). 

The big idea behind ARA is to reuse the deployment automation across the application lifecycle. automate once-run anywhere :-). You can expect to see ARA solutions integrate with ALM solutions such as Continuous Integration servers (Jenkins, TFS etc..) and others. these solutions all provide audit capabilities for compliance and ability to maintain and up to date software inventory so you always know what version is deployed where.

Thomas - I hope this helps, If you want to share more details about your specific challenges I would be happy to help more if I can.

[Gregory Damiani]

We use Puppet for config management (plus CloudFormation to orchestrate building/rebuilding stacks in AWS) and Jenkins for CI and scheduling deployments to different environments.  Fabric scripts inside Jenkins do a lot of the heavy lifting as far as getting our code onto servers.  When we come up with a new feature or bug fix, we roll it out to 6 (and growing!) different university's websites in a serial fashion.  Streamlined and Repeatable are big watch words for us.  Maybe we should be looking at Nolio or something similar.  

Perhaps it's been done on this list already, but is anyone interested in sharing their config-build-deploy stack?  I figure with enough responses, we can all get a sense of what state-of-the-art means for the group.

Greg Damiani

DevOps Engineer - 2U, Inc.

  e: gdam...@2u.com

  s: gregory.damiani

  p: (646) 597-5506

  c: (845) 548-6361

[mpron]

I interviewed Wesley Pullen, a VP at UC4, recently.  He seemed pretty passionate about the robustness of use cases that their platform has templates/recipes for.
He said they have 450+ templates for multiple job environments and operating systems.  So whether you're using 1 server or 1800 servers, the deployments and configurations will work.   UC4 keeps track of not only dependencies, but network settings and specific component details for hundreds of open source tools that might be used, like Maven, Jenkins, or Subversion.
They recently added a tight CollabNet TeamForge integration.

I have no idea how that compares to the other vendors though.  Who are all the ARA vendors?  Google gave me Attunity, XebiaLabs, and BMC along with UC4.

-Mitch
DZone Writer and Curator

[Adrian Cockcroft]

Netflix uses Jenkins and is in the middle of open sourcing our build and deploy systems. Just about all the long running processes and sequences like upgrading a Cassandra cluster one node at a time are done using Jenkins. 

Adrian

[Diptanu Choudhury]

Ditto. I came out of a project, where we used Jenkins to run all the long running processes. As long as the Jenkins Master does the work of only orchestration and all the heavy lifting work in the slaves, the model scales pretty well.

--
Thanks,
Diptanu Choudhury

Senior Consultant, ThoughtWorks
Mobile - +447808024919
Web - www.linkedin.com/in/diptanu

Twitter - @diptanu

[Ravi Chintamareddy]

Hi Thomas,

 

May I ask what implementation model/tool your team has finalized after considering these inputs/options from all our friends, please?

 

Regards

Ravi

[Thomas Falkenberg]

Hi Ravi,

we are currently focusing on Nolio. 

Regards,

Thomas