Hi,
I need to track Element counts across all our systems in 1 LCA, to start I've tried using GetElementsForView and GetElementsLiteForVIew but it's only working with viewIDs that aren't the root view '-1' which is the one I need. Is there a work around to get the result I need?
GetElementsForViewCount returns correctly with -1, but I need to count active/stopped Elements separately so need the full DMAElement array to do some extra processing.
Is there any other way I can achieve the same result. Once I have the Elements I then need a way to store the values so I can trend historical data, is that possible?
Thanks!
Ross
Hi Ross,
When you say 'all our systems' I assume that you are trying to monitor multiple DataMiner clusters. Is it correct my assumption?
How are you executing these requests? Through a ad-hoc data source?
Hi Miguel, I've managed to get it working! Yes, I should have mentioned this is across around 10 DM Clusters for a centralised view using ad hoc data source and the GetElementsLiteForView method. (Using GetElementsForView wouldn't complete, I always got a 500 internal error for systems with around 8 – 11K Elements)
The next part of the requirement might be trickier, is there a way of storing the data so I can see trends over time? Would I need to push the table I have into a DOM Instance or something?
Hi Ross,
For that use case, a possible approach is to leverage the Data Aggregator. This DxM allows you to execute GQI queries on a scheduled basis, with the results automatically exported to a CSV file.
For example, you could schedule the GQI query to run daily. Each execution will generate a separate CSV file containing the query results. From here there are two possible options:
- Using an ad-hoc data source, you could process the content of the CSV files and output the data as a time series. These results can be visualized in a line chart component.
- Using a connector, you could process the content of the CSV file and display it in a table. By enabling trending on the columns containing the aggregated values, you could visualize this data in a trend graph or a line chart within a dashboard or LCA.
Hope it helps.
Hi, I've done some more testing this morning and it appears to be working for some of our systems. The only thing standing out to me at the moment is it seems to not respond from the larger clusters. Could it be the response is too big to deserialize or is taking too long?