* Reviewing some old folders reveal some interesting facts
Some issues of this newsletter are inspired by major breaking news – an acquisition, an important product launch, or occasionally a conference. Most are built around a focused set of ideas generally inspired by some weekly dialog with a client or a colleague. Some columns – and this is probably my favorite category – are inspired by fresh EMA research: research that’s either underway and I’m looking for your involvement, or breaking news in the “data and statistics” category.
Today’s newsletter was inspired by cleaning my office. EMA is moving its Portsmouth office several blocks and of course this demands a review of some old files. Some are simply nostalgic but I had to ask myself – how much has changed and how much has remained the same? I think you’ll see some pretty clear examples of both.
I’m going to start with some fairly vintage data on application management, this one from “The ASP Industry News” from April 2000: “Although a big challenge for ASPs [application service providers], application customization may prove inevitable in the future.” This has turned out to be, as they say, a pretty safe bet.
And here’s some old research from EMA. A survey carried out in the third-quarter of 1999 showed that application managers’ priorities in monitoring for availability were: Windows NT servers (cited by 60% of respondents), Unix (45%), mainframe (31%), desktop (12%) and other (8%). Linux wasn’t around then and would have made a difference to the survey today, but Windows is still No. 1 in many enterprises.
Network connectivity for applications, on the other hand, has changed even more dramatically. In research from the second quarter of 2000, the following were preferred transport types: Dedicated T-1 or below cited by 60% of survey respondents, analog modems (45%), dedicated T-1 to T-3 (43%), ISDN and frame relay (40%), XDSL and cable modems (20%-plus). This was followed by satellite, Asynchronous Transfer Mode and other. There are some clear changes in this picture and interestingly EMA should have some fresh data here in within several weeks. But the fact that MPLS and IP-VPNs didn’t even cast a shadow (albeit this was in part due to the questioning) does reflect some pretty major changes.
Another article from 1999 talked about three major ways to monitor applications (not from an EMA analyst): client-side agents, synthetic transactions, and probe-based or flow-based information. While there have been some huge advances within each of these categories, they still pretty much apply to application performance monitoring options today.
In the failed predictions category, I came across, from multiple pundits (none EMA though) the conviction in late 1999 that Microsoft was going to be “broken up.” In the “technologies on the rise” category, an article in New York Times from April 2000, showed the percentage of employees receiving and storing e-mail had risen from 14.9% in 1997 to 38.2% in 2000. I guess it’s safe to say that the climb has continued. The article gives advice including avoiding files with .exe attachments, being wary if you receive scores of messages from the same person, and taking care not to use racial slurs or words like “sex” and “babe” in e-mails.
And then there’s ancient history – 1996. A chart about SNMP in an EMA-sponsored seminar makes the following claims:
* Nothing on the horizon is going to displace the vast installed investment in SNMP.
* Harmonious marriages between SNMP-based management and Web-based management will ensure and extend the life of SNMP-based management.
* Web-based management is going to increase the need for SNMP-based definitions, instrumentation and collection.
And if all this seems to still cast a lingering shadow of truth then how about this? An EMA chart from 1996 described the state of the enterprise management industry as: “incomplete, immature, ill-defined, evolving, and more art than science.” I leave it to you to decide how much has changed there.
