October 16, 2007 at 12:20 pm
This may seem pretty stupid, but I can't help but think everyone has ran into this issue at some point in their careers.
I just started a new position and I am trying to do some planning so I can make sure the next round of hardware i purchase will hold up for a while. The problem I am having is that I have extremely limited historical data to extrapolate. The only data I have his DB size over the past year and the only reason i have that is because I can look at the backups.
I can get the exact number of users/transactions that occur and this moment in time, but this is also a relatively low traffic time of the year.
I could use some advice as to how others have over come the problem of future planning with out historical data.
Thanks for your advice in advance!
October 16, 2007 at 12:57 pm
Capacity planning is tricky even with the right historical data. Like the argument I keep having with my real estate agent: historical data is useful only when you have some indication of how or whether it will be indicative of the future. If a new circumstance not ever observed in past history occurs - past data doesn't or won'/t know how to account for that. One constant in business is that just about nothing in business stays constant, so planning solely for "the past" still means you might not be ready.
The best place to go for feedback is your end-user community. Start by figuring out if what you already have out there is "good enough": are you meeting the responsiveness/performance they need to do their jobs? Are there areas for improvement? You might need to spend some times with your dev team to figure out which issues are SQL-based and which might be application-based. And don't be afraid to ask for direct input from the departments, and NOT from the managers themselves - you don't need the data "filtered/interpreted".
The managers on the other hand would likely have some idea of new things brewing/new initiatives/new needs. In other words - what might change. Try to use that to see how you should use your historical growth data.
And then, once you have your number, the one that matches all of the needs... Double it. That's what you're looking to go for.
----------------------------------------------------------------------------------
Your lack of planning does not constitute an emergency on my part...unless you're my manager...or a director and above...or a really loud-spoken end-user..All right - what was my emergency again?
October 16, 2007 at 3:56 pm
In addition you should talk to your "clients" and find out what "percent" increase are they expecting in their business and factor that in!
I also go for the double rule that Matt posted above but it works well when numbers are not scarily $$$
Cheers,
* Noel
Viewing 3 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply