Strange negative values in Statistics

[cclub75]

I've just started to test FMS 9.

After one hour of using my solution (20 files) with a few clients... I saw some strange negative values in the column "average" of the Statistics.

For instance :

Time waiting/call : -1779232768

Time spent/call : -1416167424

(sorry, I use the french version of FMS 9, so I don't have the exact translation on the english version)

What does it mean ?

[Don_Macaroni]

Is your system time/clock OK? Since your dealing with time, and it is negative, it seems logical your time is set in the past.

[cclub75]

I will check monday.

But "in the past" compare to... what ? There is only one clock on the server.

As for the long numbers, I found something in the doc : "In the Database Statistics table, the following units are measured in microseconds (0.000001 or 1e-6 seconds is 1 microsecond): Elapsed Time/call

I/O Time/call

Wait Time/call

So, for example, a value of 1 million means an average of 1 second per call."

It's rather strange from FileMaker ! Anyway.

But still, it doesn't explain the negative values.

Edited April 12, 2008 by Guest

[Steven H. Blackwell]

These times are in microseconds. So they should be divided by 1 million to get the time in seconds. Still doesn't account for why they are negative number though. Are you running the latest v-rev of FileMaker Server?

Steven

[cclub75]

Yes Steven, it's the 9.0v3.

I will look carefully on monday, and I will post a screenshot if I see again those negatives values.

[cclub75]

I knew I was not crazy. :

Here are 2 screen shots... with negative values.

I've checked the time, the date on the server...

As for the parameters for the stats, I was on defauts values : 15 seconds, and 40 MB for the size of log file.

Then after I changed to "1 minute" for stats... same result : some negatives values.

The server is a brand new Dell PC. With Windows XP on which I've installed the demo version of FMS 9.v3

Apparently, our solution is working fine (20 files, for 25 clients, some scripts in the night).

But still, those negative values for a time mesurement do not make sense.

Edited April 14, 2008 by Guest