Roughly a year after closing the deal, Hewlett-Packard has taken the first major steps towards intertwining the Mercury Business Availability Center offerings with its classic OpenView stack.
And in the process, HP re-architected many of the network monitoring tools that put OpenView on the map in the first place.
At the centre of the announcements, HP has tied in pieces from the former OpenView and Mercury product lines and tied them to its Universal CMDB, to provide a common picture of IT infrastructure configurations. It has tied Network Node Manager, Business Availability Center, Operations Manager, and Service Desk to the CMDB so everyone, from operations to service desk and all levels of support, literally read off the same page.
As to individual product refreshes, the highlight is the new HP Network Manager i 8.0, which will be abbreviated with the unpronounceable acronym NNMi 8. It is the rewrite of the venerable tool that initially served as the cornerstone of HP OpenView. Specifically, NNMi 8 was rewritten on a 64-bit distributed architecture and changed the node discovery mechanism. In place of the agents that NNM has long relied on, the new version instead pings the network constantly to produce more realtime, or current system topologies.
Additionally, a new root cause analysis engine was written atop the higher powered, more distributed discovery engine. It also uses new filtering mechanisms to reduce the nuisance alarms that overwhelmed traditional network monitoring consoles. In turn, a web service API will enable HP's updated network discovery, monitoring, and troubleshooting tools to more readily integrate with third party tools.
HP has also updated Operations Center, Center, the former Mercury module that served as a higher level dashboard of applications, rather than system performance. It has added tools to help automate the definition of service levels, based on data from the existing Appilog application mapping engine.
"This enables you to focus on change thresholds, and understand interplay between devices," said Ramin Sayar, senior director of BSM products.
Business Availability Center (BAC) in turn absorbs technology from last winter's Bristol Technology acquisition, which added the TransactionVision tools to help full in the picture on end-to-end performance. In fact, the absorption into Operations Center is more of an incremental step, as Bristol previously had fairly deep integration with the product long predating the HP acquisition of Mercury itself.
According to Sayar, the integration brought together BAC's end user performance monitoring with the tracking of transactional performance handled by Bristol. In so doing, the new integration resembles CA's acquisition of Wily Technologies, which also combines transaction and client side performance views.
Although not new with this release, Business Availability Center already had some links with Opsware's Policy Automation System (PAS) at the data collection level, which automates workflows for remediating service level issues.
Additionally, HP is continuing Mercury's practice of offering Business Availability Center through a Software-as-a-Service (SaaS) offering. And HP is continuing Mercury's old practice of introducing new features first through the SaaS offering prior to formal release.
The new tie-ins with Universal CMDB have therefore been available to SaaS customers for several months, and are entering general release now.
Obviously this is yet another step in the ongoing saga, as HP steadily brings more of the pieces that it has acquired together.
CMDB integration is therefore a necessary, although hardly surprising announcement. It is also the necessary first step before adding fuller process integration. Clearly HP still has its work cut out.
But what was the most interesting aspect of this release was the rewrite of HP's classic Network Node Manager. NNM helped pioneer SNMP device monitoring and for many years was the core of OpenView; if you were an OpenView customer, chances are, it was because of NNM. And so, despite HP's pre-Mark Hurd incompetence when it came to selling software, NNM was the heart of the business that the company could not kill off. It made sense for HP now to revisit this tool and not just rest on its laurels.
The tying together of assets spanning OpenView, Mercury, and Opsware open up some interesting possibilities. For instance, if you could correlate low level device performance data from NNM with the application mapping and SOA governance technology that came from Mercury could provide useful infrastructure insight as to why particular web services regularly fail to meet their SLAs. And that is just a glimpse of what could happen when you start mashing up the various pieces that HP Software is just now starting to put together.