...
For change archive site architecture from flat-files to relational databases we first make a database schema which is suitable for archive site data requirements i.e tables instead of flat-files are used for collection,processing and storage of data. Our final purposed has three tables one has nodes information, one has metadata and last table is data which contains results. Schema is shown below:
Host table containing all PingER nodes information:-
Metadata table contains all metadata of results:-
Data table contains final results of different network metrics:-
The Relationship between these tables of PingER Schema is shown below in an ERD Diagram:-
Implementation
To implement archive site according to the tables shown above we have a need to combine two scripts. One for data collection and one for hourly data analysis. So we combine getdata.pl script with analyze-hourly.pl script and formed a new getdata_new.pl script which do the same task as these both scripts collectively have done before. The requirement of combining these two scripts comes because in this new schema raw data (ping data) which is collected from different monitoring sites is processed at run time and generated results are stored in data table while in previous flat-files version raw data is stored in separate flat-files by getdata.pl and then analyzed by analyze-hourly.pl which stores results in other flat-files.
...