Hi,
I created a table which recover traffic data Uplink / Downlink with a timestamp. Is there a way to put those data in a graph with the timestamp in abscis and the traffic values in Y axes? The table looks like that:
Index [IDX],Time Stamp,Traffic Download (kbps),Traffic Upload (kbps)
1,167694480000,47.615,47.688
2,167694510000,36.680,33.714
3,167694540000,60.289,56.053
4,167694570000,46.562,46.864
5,167694600000,52.484,48.477
And also, the timestamp is in Linux ms format. Is there a way to display it in "dd-mm-yyyy hh:mm:ss" ?
Thank you,
Hi Dominique, this cannot be done directly.
In your connector you could however add Traffic Download (kbps) and Traffic Upload (kbps) non-table parameters, and use the history set functionality in a QAction to process the retrieved data from the table and perform a set on these new parameters with the correct value and time information. That way, if you trend those parameters you will be able to use all trending features available in DataMiner on those parameters.
For more information on the history set functionality, please check How to use history sets on a protocol parameter.
Hi Dominique, something like below code should work. I’ve hardcoded the unixtime and value in the example, you should replace this with the retrieved values. Also check if the retrieved values are in UTC or local time and adjust accordingly.
int value = 123;
long unixTime = 167694480000;
DateTime unixEpoch = new DateTime(1970, 1, 1, 0, 0, 0, DateTimeKind.Utc);
DateTime dt = unixEpoch.AddSeconds(unixTime);
protocol.SetParameter(Parameter.yourParameter, value, dt);
Well noted. I will try in the next version, thank you 🙂
In case it's only about presenting the data: In a DataMiner dashboard or Low code app, you can present any data you can access through Queries in a line/pie/donut chart. (Dashboards or low code apps can be embedded in Visio files in case you want to have them embedded in Cube).
So it's just a matter of creating the correct query in your app/dashboard. You can easily get to your parameter table I assume and then you have to split the data in different columns via column manipulation operators.
Thank you Michiel,
So what I did, I took the most recent record of my table which is polled every 5 minutes and put it in a standalone paramter. That way it works indeed.
What about the linux time stamp? I can’t fine a data type to convert it in “human readable” format. I wanted to create a sub function in the QAction which recovers the table but I got a message saying that I need to upgrade my C# version. Any suggestion?