jason w. jones

software from the trenches

Making order out of chaos with Essbase Integration Server – part 2

Last time I gave sort of a high-level look at pounding the rigid financial page definitions from the corporate overlords upstream folks, and generating the outlines based on this information. So now I have some nice fresh outlines, but I don’t have any processes for data just yet. I have a couple of issues I need to address:

Continue reading Making order out of chaos with Essbase Integration Server – part 2 →

Making order out of chaos with Essbase Integration Server – part 1

Essbase has been a great addition to the decision-support toolbox in our enterprise due to the fact that the financial pages are produced via a “home-cooked” solution. It’s actually quite impressive, although in some ways I would say that it is limited due to it’s rather rigid structure. Our world consists of seven-digit accounts for everything. You can glean a little bit about what the account is based on what digits are where, although don’t count on it. Accounts are then associated with lines (NET SALES, GROSS PROFIT, RENT, EBITDA, etc), and the lines make up the “physical pages.” The lines are numbered 00 through 99 and they quite literally correspond to a position on a piece of paper. 00 is the first line, followed by 01, then all the way to the next page with 99 on it. It’s not necessary to use all the lines, in fact, many of them are blank. Accounts are then directed towards lines. They are not “nested” in a true hierarchy — an account is simply directed multiple times. For example, sales for a particular department are directed to the NET SALES line, but since it also affects GROSS PROFIT, that particular account is also directed to the GROSS PROFIT line. This is all maintained in the Cross Reference, which is the master list of all report types and account groups (basically just different accounting systems for different business segments), the account, the line it rolls to, a description of the account, and a description of the line.

Continue reading Making order out of chaos with Essbase Integration Server – part 1 →

A day spent on tuning

I spent most of the day working on the test server. Right now it takes about an hour, start to finish, to load up a new period in the period financial databases. There are 18 databases that need to get loaded, as well as one staging database that is used to feed these 18 databases. Currently the process takes about an hour. It’s not bad, but lately I’ve been thinking about how the process could be improved and sped up. It’s not that taking an hour is a bad thing, when the system works and all the numbers tie out, because we don’t generally need to process data more than once an hour, but if the numbers don’t tie out for some reason, then it’s time to reprocess, and there goes another hour. So how does the current process work?

Continue reading A day spent on tuning →