Automated event management is the future of systems management packages like Computer Associates’ Unicenter TNG. Susan Amos from our sister publication, Software Futures looks at what AEM practitioner Maxm has to offer.
Charles Wang, chief executive of Computer Associates, has high hopes for the company’s 3D systems management product Unicenter TNG, slated to hit the streets next month. He intends his new toy to eventually look after not just your regular Unix systems and mainframes, but also monitor air conditioning units, lighting systems and electronic doors. Now meet one small company that’s beaten CA to it. McLean, Virginia-based Maxm Systems Corporation. It’s the systems management company you’ve never heard of, but counts among its customers the cream of US banking and telecommunications companies. To be a little more specific, Maxm’s impressive customer list includes Bank of America and AT&T. And the privately held company is obviously doing something right as its turnover has rocketed up to $20m a year, from a lowly $1.4m four years ago when it was a humble systems integrator with a toolkit. Maxm’s software Max/Enterprise does automated event management. We can manage anything that creates an event log. Everything from routers to Private Automatic Branch Exchanges (PABXs). We focus on the automated correction of faults, something that Unicenter doesn’t do, claims the company’s vice president of global marketing Pam Casale, based in McLean.
We spoke to one of Maxm’s smaller users, First National Bank of Maryland. The bank does $11bn worth of business each year. It is owned by Allied Irish Bank, Ireland’s largest bank. First National needed an overall management package, due to the mushrooming number of non-mainframe computers. That is, Tandem, Stratus, HP-UX and AIX boxes. The reason we got Maxm was not because of the mainframe, that was already under control. We brought that into the puzzle last, explains John Lewis, vice president of the automated operation group at the bank. With Maxm, Lewis has reduced the number of swivel chairs in his operations center. To manage our IT infrastructure we had a command center with 35 terminals, and needed to keep it staffed 24 hours a day. We’ve reduced our staff numbers down to half of that . One feature that’s cut down the workload is the ability to filter out unimportant ‘status okay’ type messages that sapped staff attention. Now you’re suppressing 90% of them, and massaging the more important ones, giving them more priority. We quizzed Lewis as to how he’d heard of Maxm in the first place. He understood our concern at the company’s lack of profile in the market. I’ve always said to their marketing department that they haven’t done a good job. It turns out market research organization the Gartner Group provided a list of companies that do enterprise wide fault management, and Maxm was on it. The bank then whittled its shortlist down to Maxm’s MAX/Enterprise and Boole and Babbage’s Command/Post. Lewis said it was primarily Maxm’s architecture, based, as it is, on IBM’s AIX, and its ability to run OS/2 clients that swung it for him. However another Maxm customer, European IT services company Cap Gemini, which has the largest FM capability in Europe, servicing government departments and utilities, was more critical of rival Boole and Babbage. The main reason Boole and Babbage fell off Cap Gemini’s shortlist was failure to return phone calls.
Returning to First National’s experiences with Max/Enterprise, Lewis adds, Also the software was more robust in its ability to automate the actions you need to take when an alarm is triggered. He finds the package very customizable: You can let your imagination run wild tailoring it to your organization. Using Maxm Lewis has built what he calls a notification services hub, which can send a warning out by Lotus cc:Mail, the bank’s internal email system. If it’s a humdinger of an error, Maxm can page the person with the best chance of solving the problem. We page the right person with the right tool. It cuts out the middleman. Because Maxm is very flexible, it doesn’t matter what pager system or trouble ticketing you’re using. The bank uses IBM’s InfoMan, and has woven it together with Maxm, so problem details are entered automatically into a trouble ticket. Lewis has also set up a rolling message board for the back office. It tells staff of any service disruption and informs them when service will be resumed.
Comparing it to other products, Lewis says Maxm does one thing well, which is fault management, and covers the whole spectrum of devices. Hewlett-Packard’s competitive HP OpenView product can only manage SNMP devices, which only make up ten percent of his infrastructure. He thinks CA’s Unicenter does a whole raft of functions like software distribution and asset management but can’t yet talk to as wide a range of devices as Maxm. Think of Maxm as a master integrator for products like H-P OpenView and CA Unicenter, is Lewis’ advice. So while you wait for Charles Wang to get his new toy monitoring anything that bleeps, Maxm is well worth investigating.
