We are getting a lot of requests for reports on member transaction data, but the datasets are huge, currently about 89 million rows and we're running into major performance issues when trying to build a report or publish to Tableau Server. One of our devs suggested building individual tables at different aggregation levels to connect to, but I'd really prefer the flexibility of having transaction level data.
Wondering how others are dealing with this at their credit unions.
Thanks!
Dana
6 respostas
@Credit Unions Hi,
I guess the bottleneck if the amount of data to be retrieved from data source. In order to speed up that you can do following:
- Look for option to use in memory caching of database, for example if you are using oracle database, you can use in memory option and cache entire table in memory. This will avoid disk reads and reports will be faster.
- Use extracts which will avoid retrieving huge datasets (this would obviously depends on how much stale data you can bear in your reports).
