Quest Software holds frequent customer briefing events all over the world – where we talk with our customers about their challenges, explain our own thinking and hopefully agree on mutually-beneficial areas to explore.
The most recent took place in beautiful Amsterdam and Hamburg last week. Prospects and customers from all over Europe and the Middle East came together to discuss application/database management and Windows infrastructure management. Their profile was varied – representing finance, manufacturing, telcoms, retail and transport.
So what were the ‘hot topics’ in applications & database management?
Firstly, virtualisation: everyone’s talking about it, a lot of people are doing it, but not so many people have yet thought about what virtualisation does to application performance management. What it does it this – it breaks (even further) the ‘traditional’ view of application monitoring which ties monitoring to particular physical assets (usually a server). In a virtualised world, a particular app can be served up from multiple virtual servers which are undergoing dynamic (re-)provisioning. Each of those virtual servers takes on part of the resources of any physical server. So what does “CPU is at 100%” mean in a virtualised environment? Real CPU? CPU allocation for a particular VM? Something else? And if the virtualised app servers are moving around the real servers then *does it matter* if that particular physical server has 100% CPU? The only way to know is to be able to map the real, changing physical environment to a logical model of your environment that tells your management toolset the meaning of particular alerts. As virtualisation spreads, I’d go so far as to say that asset oriented performance monitoring will die, because it’ll become meaningless.
Secondly, collaboration in application management. Loads of our attendees told us that they are in, or are thinking about, outsourcing agreements or other situations where their environment will be managed by multiple teams, perhaps in different places. In all cases, they already have management toolsets, but those toolsets simply don’t help different teams collaborate who have totally different skill sets and who have completely different preferred ways of visualising performance data. Particularly interesting were the ways in which SLAs can get mangled between intent (save money and get a better service) to a distributed non-ownership of cross-team performance problems that helps no-one. Just like herding cats. Much of this is about management processes and contracts, not technology, but a ton of it spills over into performance management where people want to have ways to share problem diagnosis.
Thirdly, our customers wanted to talk about how they could get a better understanding of how their real end-users are experiencing the application. A minority of our attendees were already using End-user performance management tools, but there was a lot of interest in using them more – and in tying them into the rest of the application performance data that they are already collecting. About 2/3rd of attendees reckoned they’d be explicitly managing end-user performance over the next 12-18 months. That represents a lot of potential growth in a market which, for example, Gartner regards as having moved some way out of the trough of disillusionment over the past 12 months.
So, despite the very varied attendee organisations, we’ve seen yet again that people are re-evaluating their performance management toolsets in the light of recent market changes. We were happy to see that our attendees broadly validate our strategy of enabling users to more easily adapt to change, visualise varied performance data in multiple ways and collaborate more easily with colleagues in other organisations.
Would love to hear if you’re going through the same thought processes.
Hugh McEvoy
Manager, PM
Quest Software, Inc.