Anyone here use BMC's tools for DevOps

[Pierre Boudreau]

Hi,

Does anyone here use BMC's Application and Middleware Automation tools ( http://www.bmc.com/products/offering/application-and-middleware-automation-devops.html ) ?  I'd like to know how it compares to something like OpsCode Chef with regards to functionality and pricing.  I haven't found anything very concrete on their website.  I only have a couple days to make a decision and I'd like to get a sense if it's worth persuing as an option.

Particularly, I have a prospective client who wants us to use (or integrate with) their BMC CMDB product as part of a solution.  If I want to implemented a deployment pipeline that automatically keeps their CMDB up to date, would my best bet be to go with a complete BMC solution, given this constraint?  Would this be achievable by integrating other tools like chef-and Jenkins with BMC's CMDB?  Has anyone done this sort of thing?

Thanks,

Pierre

[Jim Hopp]

We used BladeLogic at a previous employer. We had a few hundred nodes at the time and used BladeLogic for application deployment. It did what it was supposed to do, but it was quite expensive and was overly complicated. BMC was quite keen for us to use the rest of their suite (CMDB, etc) but you really have to buy into their whole stack. We were just starting to deploy Chef when I left; I believe they still (a year later) use BL for app deployment but Chef for machine configuration.

My belief is that the industry is moving quickly to open-source tools that take the Unix approach (use simple tools and build complex processing out of them) and pay for services (e.g., hosted Chef) if desired. The days of complicated, expensive, single-stacks solutions like BMC are, IMHO, numbered.

I've only used open-source Chef so I can't compare BMC pricing to Opscode pricing.

-Jim Hopp

[kimi...@gmail.com]

are you still interested in data?  I am doing a large devops deployment from application to infrastructure using some BMC.

Kimizu

[Mark Jaffe]

My organization bought an enterprise-level installation of Varalogix-Q, which was bought by BMC this past August. I was somewhat underwhelmed by their proof-of-concept which was presented last April, since I had already been evaluating Nolio for my own department. But the decision made by others, since a lot of the automation was going to be done on Windows platforms. I did automate a few of my deployments but abandoned the use of Varalogix for my deployments due to unpredictable behavior. Support was pretty good from the original vendor, but BMC seems to not be too interested in providing support. I cannot speak for any other of their products.

Mark

-- 

Mark Jaffe

jaffe...@gmail.com 

408-807-2093 (cell) 

408-341-9263 (main)

[Stuart Charlton]

Old thread is old but I'll chime in.

The varilogix product has come a long way since last year, with quite a few more integration points and ability to debug your Varilogix plan prior to deployment which makes it much more predictable.

The overall devops product is "Release Lifecycle manager" which is the combination of StreamStep's release workflow and Varilogix for deployment orchestration.  There is integration now with Bladelogic and CLM for provisioning, along with Remedy for change integration.  Not aware of OOTB Atrium CMDB integration yet.

(I'm a new BMC employee but don't speak for them)

Stu

Sent from my iPhone

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Matthew Simon]

I used some of the BMC tools at my current company, and dumped them.

Background: The infrastructure that I deploy to is 70% windows server 2008, 20% RHEL, 10% Solaris, about 1200 nodes total. We had a 'automated' deployment system built on top of Bladelogic, but frankly it was nothing but trouble. False successes were the norm, as were reliability issues, and lack of ability to paralleled tasks in the ways I needed. Also - the cost for just his product is _extremely_ high in my opinion, and to get the most value out of the product, you need to buy into a lot more of the stack - CMDB, Remedy, Runbook Automation, etc. This can run into the millions of dollars for a shop my size.

I looked at many options for replacing this, including Chef, Cfengine Nova, and Puppet Enterprise, and did trial implementations of CFEngine and Puppet Enterprise. The most constraining factor for me is the need for robust windows support; which rules out a lot of open source tools, and frankly CFengine and Puppet were extremely rough around the edges with windows - far too rough to use.

Ultimately I decided to roll-my-own based on some open source components; Jenkins as the main trigger, SVN doing file syndication to any server that needs packages libraries, Nant + Lots of Powershell + Wix for the windows targets, Ant + Bash + RPM for the linux targets, and for targets were I don't need to actual get on the end system to action them (Like Oracle or Control-M), I touch them remotely from a windows jobs via SQL plus or similar.

Its definitely missing some of the things the off-the-shelf products have, but I am very happy with my choice, because I have a team (of 3), that knows the system like that back of their hand, can address bugs immediately, and is constantly adding the exact new features that will give us the most value.

In terms of cost comparison - I takes me about 2x as much to fund this team than I was paying for Bladelogic maintenance (NOT including BL licenses), so I'm pretty comfortable I'm getting a good deal here.

-Matt

[Randy Ausenhus]

We are looking at doing the same thing right now and BMC is one of our options.  Could you provide me with some feedback around your experience?

 

Thanks,

Randy

[Mark Jaffe]

Do not do it, we have had nothing but  headaches from that tool (specifically what was formerly known as Varalogix-Q)

Sent from my iPhone

--

You received this message because you are subscribed to the Google Groups "devops-toolchain" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Clyde Jones]

It was expensive, difficult to use, and didn't cover all our use cases. For the most part, it was more difficult to use bmc than to do it all by hand. 

--

You received this message because you are subscribed to the Google Groups "devops-toolchain" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
We are what we think. All that we are arises with our thoughts. With our thoughts, we make the world.
-Buddha

[Ajey Gore]

User Chef or Puppet or Ansible, way less headache.

Thanks

Ajey

[Tim Brown]

Agreed.  I'd add Salt to that list, but I'm currently favoring Ansible.

We used it to replace 9 months work of CA Lisa release automation (Nolio) in 6 weeks

If BMC is similar I'd run run run.

If I -had- to use a pointy-clicky tool it would be Xebia's.

[Josh Pederson]

Curious... what prompted your move from Nolio to Ansible?

 

On Sunday, October 12, 2014 4:50:34 PM UTC-7, Tim Brown wrote:

Agreed.  I'd add Salt to that list, but I'm currently favoring Ansible.

We used it to replace 9 months work of CA Lisa release automation (Nolio) in 6 weeks

If BMC is similar I'd run run run.

If I -had- to use a pointy-clicky tool it would be Xebia's.

On October 12, 2014 at 10:05:39 AM PDT, Ajey Gore <ajey...@gmail.com> wrote:

User Chef or Puppet or Ansible, way less headache.

On Tue, Oct 7, 2014 at 9:01 AM, Clyde Jones <slash5...@gmail.com> wrote:

It was expensive, difficult to use, and didn't cover all our use cases. For the most part, it was more difficult to use bmc than to do it all by hand. 

On Monday, October 6, 2014, Randy Ausenhus <randy...@gmail.com> wrote:

We are looking at doing the same thing right now and BMC is one of our options.  Could you provide me with some feedback around your experience?

 

Thanks,

Randy

On Wednesday, February 13, 2013 9:58:45 PM UTC-5, kimi...@gmail.com wrote:

are you still interested in data?  I am doing a large devops deployment from application to infrastructure using some BMC.

Kimizu

On Sunday, November 25, 2012 10:53:10 AM UTC-5, Pierre Boudreau wrote:

Hi,

Does anyone here use BMC's Application and Middleware Automation tools ( http://www.bmc.com/products/offering/application-and-middleware-automation-devops.html ) ?  I'd like to know how it compares to something like OpsCode Chef with regards to functionality and pricing.  I haven't found anything very concrete on their website.  I only have a couple days to make a decision and I'd like to get a sense if it's worth persuing as an option.

Particularly, I have a prospective client who wants us to use (or integrate with) their BMC CMDB product as part of a solution.  If I want to implemented a deployment pipeline that automatically keeps their CMDB up to date, would my best bet be to go with a complete BMC solution, given this constraint?  Would this be achievable by integrating other tools like chef-and Jenkins with BMC's CMDB?  Has anyone done this sort of thing?

Thanks,

Pierre

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.

To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolchain+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
We are what we think. All that we are arises with our thoughts. With our thoughts, we make the world.
-Buddha

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
Thanks

Ajey

[Stuart Charlton]

For what it's worth, as a current BMC employee speaking for myself

- BMC has retired Varalogix and sells Release Process Management (RPM / the former StreamStep) optionally bundled with BladeLogic Server (the enterprisey precursor to much of Puppet/Chef/etc).  Varalogix is still supported and will be for a few years.

- There's BladeLogic Middleware Automation (the former Phurnace).  This is mostly for deep snapshot/compare/config of J2EE app domains or cells, which has proven (to me) helpful part of a nutritious devops toolchain if you have a lack of scripting discipline

Whether you can do it with Chef/Salt/Ansible or BMC (or CA for that matter) depends a lot on your team and tech base.  If you want to IIS/.NET for example, or old school WebSphere, or have a team that isn't comfortable with open source, you might want to look at the expensive commercial offerings.   Or a mix - it's a tool chain most orgs need after all, not one silver bullet product.  

Again, just my two cents.

Stu

Ps., I loved Varalogix Q, one of the reasons I joined BMC in the first place, and we've got some customers doing amazing things with it, but times change.  They're merging the capabilities with RPM and Blade.

Cheers

Stu

Sent from my iPhone

[Tim Brown]

There were a variety of factors:

- No ability to 'version' a deployment script.  Imagine you have script "DeployAppA" that you used to deploy to Production.  Now back in Dev we're changing  few thing with AppA and that requires changes to "DeployAppA".  Any changes you make to DeployAppA will also be used for production.  You have to basically copy the script to a new name to ensure there are no conflicts.

- Performance of artifact copies were horrid - 5x slower than ssh/scp.  This might be a specific use case  -  artifacts are produced in an EU dev center, production is a US data center, connected via WAN.  An scp of a typical artifact might take 2 minutes, but Nolio's copy would take 10 minutes easily.

- The pointy-clicky UI can add a lot of latency, and the overall use of it was cumbersome. (disclaimer: the people using it are your typical command-line *nix types.)

- We hit a number of bugs (I'm forgetting the specifics of them now) and while patches do get issued, they don't get issued fast enough if you're on the critical path.  We had a new data center launching in 12 weeks, and needed to have all the infrastructure & app stack wired up and working within 2-3 days.  

Ansible was something of a backup strategy.  A small team of people started the Ansible approach (with support from VP Ops) after the above factors were really putting some fear in our dates.   The Ansible team rocked it, and the backup became the primary.

HTH,

~Tim