Hi Henrique, We'd the same access restrictions on the Linux boxes to be monitored and had to adopt the same approach. Run a shell script to monitor VMSTAT - Run the test in Silk Perfromer - After the test get the Shell Script's log - Run it through CSV2TSD - Import the TSD in Perfromance Explorer Some learning: 1) All column values in the CSV should be double quote enclosed... IN general CSV needs double quotes only if there is a comma in the data column. But Silk's CSV2TSD needs each single column value enclosed by double quotes. 2) The CSV header line should follow the format given in the help. The help manual provides this "(PDH-CSV 4.0) (W. Europe Standard Time)(-60)","\\JONES\Processor(_Total)\% Interrupt Time","\\JONES\Processor(_Total)\% Privileged Time","\\JONES\Processor(_Total)\% Processor Time" - The first column of the header is supposed to indicate timezone, but per my analysis, it is irrelevant. See 3) - The subsequent columns can have the names of the readings. You can have backslash separated values as column names. It is used by Performance Explorer to construct a hierarchical tree of your metrics. For example if you have 2 machines and you need CPU and Memory metric of each, you could have the something like these... "\\Machine 1\CPU Utilization %", "\\Machine 1\Free Memory (MB)", "\\Machine 2\CPU Utilization %", "\\Machine 2\Free Memory (MB)" - The first column of each row should have the timestamp in the following format MM/DD/YYYY HH24:mm:SS.000 ... The last 000 is milliseconds. I don't need that granularity so I default to 000. 3) The last important piece is which machine you run the CSV2TSD tool. You have to run the CSV2TSD tool on a machine having same timezone as the machine where readings were taken. The tool completely ignores the timezone mentioned in the header (AFAIK) and uses the timezone of the machine where it is running. It reads each timestamp in each row and converts it into GMT timezone and stores it into the TSD. So that it can be converted to any timezone depending on which machine Performance Explorer is used to open the TSD. If you still have issues with the CSV2TSD tool, please provide some more details.
↧