Comments on: An Open Source CMDB

By: dlog

dlog — Thu, 29 Oct 2009 09:04:54 +0000

You should add i-doit.org to the existing ones as it is currently the best open source CMDB

By: rg

rg — Thu, 03 Jul 2008 05:42:19 +0000

http://www.i-doit.org Big CMDB-Tool written in PHP

By: jinxfei

jinxfei — Sun, 09 Dec 2007 03:10:28 +0000

htt://www.onecmdb.org
This is a simple cmdb writen in java.

By: cote

cote — Fri, 07 Dec 2007 15:33:18 +0000

Thanks for the re-reply, Marv. Ideally, what I'd like to see in an open source ref. impl. is a project that coincided with the development of the spec as much as possible. Why during? Because I'd rather see coders encounter errors in coding to the spec and feed that back into the spec as much as possible. Also, it'd be good to have the whole process of developing the code in the open for others to look at when doing their own implementation - as you nicely reference back to the SNMP impl of yore. Ideally, the ref. impl. would be production ready - if not in the fine print, then in the code at least. Also, it'd be key for it to be very permissively licensed, under the BSD, Apache, or like so that people could yank the code to use on their own or start for their own projects, open or closed source. That's just a quick, off the top of my head response. I'm of course happy to think out-loud more ;) Additionally, you might consider coming down to barcampESM this Jan in Austin, TX to talk with a wider audience.

By: Zenoss Blog » Zenoss Newsletter - November 2007

Zenoss Blog » Zenoss Newsletter - November 2007 — Tue, 04 Dec 2007 18:54:14 +0000

[...] People over Process: An Open Source CMDB [...]

By: Marv Waschke

Marv Waschke — Thu, 29 Nov 2007 19:01:47 +0000

It is too early to give you a solid answer to your question, but let me point out that for a vendor like CA, working on a standard is an investment and we will do what we can to get a return on that investment. If we don’t adopt it ourselves, our ROI is gone. If the CMDBf specification is not widely adopted, our ROI is gone.

Would an open source reference implementation help? Thinking back on my own experience as a developer, without the open source reference implementation of the SNMP trap code, (was it someone from Stanford who wrote that?) I am sure I would not have been nearly as eager to put automatically generated incidents into the service desk product I was working on at the time, and I don’t think SNMP would have become the ubiquitous standard it is today.

So, would I like to see an open source reference implementation of the CMDBf spec? You are darn tootin’ I would. I can’t go beyond my own opinion, but you know for yourself that no one likes to lose ROI…

Along those lines, tell us more about what you would like to see in a reference implementation. What form would you like it to take? How would you like to see it published? Never can tell who might pay attention.
Marv
http://community.ca.com/blogs/itservice

By: Cote'

Cote' — Thu, 29 Nov 2007 00:24:45 +0000

Thanks for the comment and link (you can bet I’ll subscribe), Marv. I agree with you: the standards are the key. My thinking is that having shipping code with the standard – rather than just having the standard itself – is vital for adoption, esp. in the IT management area. That said, I don’t know what the CMDBf’s plans for the reference implementations or other “running code” along those lines. Out of curiosity: are there some?

By: Marv Waschke

Marv Waschke — Wed, 28 Nov 2007 11:48:59 +0000

I am a partisan in this discussion, being from CA and a member of the CMDB Federation Working Group technical committee (we do have a name!), but I want to point out that the CMDBf spec and work on language standards like SML pave the way toward a level of standardization that will make an open source CMDB make sense. Believe it or not, I think an open source CMDB is a great idea, but you point out many problems that hold back an open source source implementation as much as closed source. Without widely accepted models of interfaces, CIs, and transactions, an open source CMDB is no easier to integrate or federate than any proprietary product. So right now, those of us from the vendors who are slogging it out in the trenches, building the standards that will make CMDBs easier to implement are doing a service to open source as well as proprietary products. You can see some of my perspective on this in my blog, http://community.ca.com/blogs/itservice. Marv Waschke

By: An Open Source CMDB at John M Willis

An Open Source CMDB at John M Willis — Tue, 27 Nov 2007 21:49:56 +0000

[...] An Open Source CMDB [...]