Paglo Briefing – Search over your IT Junk, and hosted/SaaS too

Earlier this week I talked with Brian de Haaff and Chris Waters about their company, Paglo who somewhat dubiously calls itself “the world’s first search engine for IT.” I say somewhat dubious because, of course, there’s Splunk who describes itself as:

Splunk is the IT Search Engine that indexes and lets you search, navigate, alert, and report on IT data from any application, server, or network device. Securely access logs, configurations, scripts and code, message, traps and alerts, activity reports, stack traces and metrics across thousands of components, from one place, all in real time.

It’s always dangerous to make claims involving the phrase “world’s first”…but let’s move on to the actual Paglo product itself.

IT Management Search, hosted

What is unique to Paglo is that it’s a free, hosted service for searching over your IT junk. In seeing the pre-recoreded demo, it really is a Goolge-like interface to simply searching around for IT things. So, you can search for “office” and it’ll show you all the Microsoft Office installs on your network. Check out this demo.

For the IT management wonks out there, there’s also some interesting categorization and taxonomy work going on behind the scenes. Check out the docs on the database, a sample database, and the Paglo taxonomy. Also, as we get into below, the crawler agent is open source; so there’s potentially plenty of dissecting to be had.

Paglo finds the aforementioned “IT junk” (or “assets” if you prefer) via a crawler agent it installs behind your firewall. That is, to get hooked-up into to the Paglo service, you have to install a program on your local network, the crawler agent. This agent scans the network at some interval and then sends up the data it finds to the Paglo service.

Thus, Paglo is a SaaS IT Management offering – a rare find now-a-days, along with FiveRuns, Service-now.com, NetCraft’s Versiera, and BMC’s Service Desk Express as the only ones that immediately spring to my mind…RIP Klir?

Coming with this hosted offering are promises of security: encryption between the agents and the Paglo service and keeping each customer/user’s data separated. While I didn’t try to hack into it or anything, it seems like they have the bases covered here: I’m no security expert, however.

What Paglo Does

At the moment, Paglo does three main things:

• Searching – searching over all sorts of IT assets and allowing users to do simple key-work queries or use the more powerful SQL-like Paglo Query Language, or “pickle” as they call it. I’m not sure what IT assets it can and can’t search – I couldn’t find a master list on their page and neglected to ask. But, as an example: we walked through a pre-recorded scenario where an IT admin found a rouge install of Office 2007, did another search to find the owner of that machine, and then did another search (on Active Directory) to find the email address of that owner to ask them WTF on the 2007 install.
• Reporting – Paglo can give you quantitate reports on things search results instead of just raw results. For example, you can see a little report with graphic about how many and what types of Office installs are out there.
• Community – Paglo is chomping on the Collaborative IT Management bit from the get-go. Users can save and share the search queries they’ve come up with the Paglo community. Thus, if someone comes up for a really great way to find Software XYZ instances that need to be patched, they could share that and let other users. I like term “social solving” term they’ve come up with the describe this.

There’s also a topology, below, that I’d have to play around with more first-hand before I commented on it.

Also of note is that the crawler runs on Windows only. Sure, it can scan non-Windows boxes, but you need a Windows instance to run it. This is pretty typical of SMB-targeted IT management offerings.

Lack of Management

Now, what Paglo doesn’t do is chain together these searches (recall the manual need to chain together who installed that Office 2007 instance) or provide any way to take action, or perform “management.” That is, if you found, for example, a user with an out-of-date password, you might like to “click here to force user to re-set password.”

Paglo doesn’t do such stuff. When I asked them about this, they gave the well practiced “we’re really focusing on delivering the best search experience for our users now” answer.

The Freeness & Open Source

The other notable thing about Paglo is that it’ll be free for users. The crawler agent is even (going to be) open sourced (under GPL). They’ll be following a Spiceworks model, then, of looking to sell ads and otherwise make cash off the attention pool of admin types.

Of course, as I suggested to them, this means it’d be wise to keep an eye on the money-making methods – “business models” – that Spiceworks has gone through over the past 2 years. That is, Spiceworks is the trail-blazer in this area, they’re successful from a user growth perspective, so it’s worth trying to duplicate and improve on their models.

The open source angle is fun, of course, as well. Once Paglo is fully out, I’ll be curious to see what the overall IT management open source world thinks of it – they’re off to a start with the very thorough docs page, e.g., notes on writing crawler plugins. Also, as I noted, since it’s under GPL, it’s worth poking around in the Linux world to see if they can get some distro-love. That’s a long row to hoe, but worth the ubiquity if it can be plowed.

Overall, I really liked the spirit of openness that Brian and Chris seemed to have. While there’s plenty of marketing-juice to be had from open source, they also seemed genuinely interested in their crawler being open source. The more thorough-than-usual amount of docs in that area are further warm-and-fuzzies.

Private Beta

Paglo is just about the enter private beta (invite only) so all of this is not really available at the moment for the general public. I haven’t checked it out myself – if I get an invite, I’ll report more.

Disclaimer: Splunk is a client, as are FiveRuns, BMC, and Spiceworks.

Technorati Tags: bmc, fiveruns, gpl, hosted, klir, paglo, saas, search, splunk, versiera