We are currently using SelectServer XM. Apparently, the newer versions offer much better reports, so I would like to upgrade my SelectServer to the current version.
TIA
When trying to analyze usage, I am having trouble counting one users usage as 1 usage. If they exit one program before opening another, I get 1 usage. But if they open a second session of the same or a different version, I get 2 usages.
I'm using Excel and cannot figure out the test or formula needed to get a true count.
Charles (Chuck) Rheault CADD Manager
MDOT State Highway Administration
Well it is a bit out of the post original subject, if you want to discuss further, please open a new one,
I guess that you've got a usage by hour reports and try to check that SELECTserver is counting peaks properly.
First thing to remember is that if you have usages that started before and ended after the period indicated to create the "by hour report", they might not be in the report but are counted in the peak usage, for example checked out license or machine that stay open a few days for some processing.
For most Bentley Product, usage is counted by machine, not by user or version.
I've done in the past an Excel sheet that do the peak analyse, I create a column by hour after the report it self then for each usage line of each hour column, if the usage finish after the hour AND start before the hour +1, then I copy the name of the machine in the cell. In a line, I count the number of different machine name in each hour column...
I’ve done that a good few times, not that I don’t trust SELECTserver developers I only love playing with Excel, and I’ve never found a difference with the peak usage indication.
Thanks for sharing that excel process - a bit over my head. What I did was first break down how many time "buckets" I was going to examine per day. If the interval is 60 minutes, I have 24. If 30 minutes, 48; if 4 minutes, 360.
I then looked at each record's start and end times to determine in which interval it started and ended. I then incremented the counter for each of those buckets (in that particular day) by one. The program keeps reading the records (previously sorted in ascending order by the start time) and figures when it crosses into the next day (20 days max, just to keep the size of the array down). Repeat.... So, with the interval at 60, i t approximates the Bentley model - I say approximates, since the program strictly looks at concurrent use in each time interval, not at how many licenses were used within that interval. As you would expect, as the interval gets shorter, the number of concurrent uses decreases.
I have been able to get Access to connect through ODBC to the Select Server. And I have found some willing guru's on a few of the Access Forums out there to assist me in building queries. I have one hour bucket queries that match Bentley's results. I am working on getting an hour by hour breakout per products to see when the parks are occurring. This may help me figure our a strategy to reduce my overages when they occur.