I’ve had several briefings recently that I haven’t mentioned hide-nor-hair about. Here’s a quick briefing 3 of those briefings, in reverse chronological order:
Opalis
Yesterday James and I had a demo of Opalis’ modeling software. The GUI is a console that lets you define workflows for IT management: check the state of a system, if it’s bad, open a help desk ticket, notify this system, then do some other action. The idea is to create a “run book,” or a set of modeled courses of action to take when some even occurs: be it reactive like responding problems in IT, or less “fire fighting” like provisioning a new server in response to a change request.
What I liked about what I saw was that it seemed to be an integration layer — a “spanning layer,” even — on-top of all sorts of existing IT management software.
Now, I’m always a bit suspicious of the wide applicability of anything that involves drag-and-drop modeling. The simple tasks you see in demos can’t speak for the long-term functionality of such a product. But, the idea and the implementation I saw looked nice. The modeling of workflows and execution of thereof certainly fits in well with enterprise ideas of systems management.
The next higher-level question is a cultural one: will sys admins initially model the various workflows they do, then maintain them? More importantly, how many workflows can be modeled an automated? One would hope the answer was many, but translating human action into programatic models can go weird. This is a train of thought that makes me wonder what applying Agile software development principals to IT management would look like.
Still, you have to start with answering the question: what is it exactly that we do? As I’ve said before, ITIL’s Service Support was an exciting, sort of fire-hose answer to that question. The key to IT getting value from commercial offering — in software or services — is parring down that fire-hose into something that can be used instead of book-shelved.
This Monday James and I talked with several folks from Curl. To use one of Bruce Sterling’s favorite metaphors, the RIA platforms are pouring out of the cracks now-a-days. Curl provides a runtime in the form of a browser plugin that compiles the Curl delivered code. Developers use the Curl language (see samples here in zip files).
As with many such technologies — complete, proprietary development stacks — it’s turning out, Curl told us that they’re big in Japan. I’m not quite sure what the connection is, but it’s worth paying attention to.
The key thing for Curl, as with any development technology, in building and maintaining a strong developer community. There’s so much duplication of technology out there, much of it for free, that a thriving community is often the key-stone to platform success.
Hopefully, you’re comparing and contrasting the general idea and concerns with Apollo.)
Greenplum
Last week, all three of us talked with RedMonk client Greenplum. The topic was an overview with an extended ecosystem discussion at the tail-end, a sort of classic RedMonk briefing/call ;>
Greenplum is another data company that builds on-top of PostgresSQL to provider enterprisey software, in this case, really big data warehouses. What’s “really big”? As one data-sheet says, “Greenplum Database scales effortlessly from hundreds of gigabytes to multiple terabytes and beyond.”
Greenplum’s strategy is classic open source and Moore’s law: because they can build on-top of the core of PostgresSQL and run on cheaper storage/hardware, they say, they can offer it cheaper than the incumbents.
What intrigues me about “business intelligence,” though, is not so much the nuts-and-bolts so much as the opportunities that lower price-points offer to previously under- and un-servered markets. The story of Wal-mart vs. K-Mart in the 80’s is one of Wal-mart figuring out how to deal with a profit from piles of retail data. I’m hopeful that smaller businesses, given access to “business intelligence” could find similar success. Of course, it could be a case of “everyone’s got it” as well…in which case the point is on figuring out new ways to use that data at an ever shrinking frequency. Sounds exciting!
Previous Topic
