I have an extract that I have been refreshing every week for over a year, with roughly ~30million rows and maybe 150 columns combined including a primary Teradata connection with two left joins to Excel documents. This has taken roughly 1.5 hours to create and has consistently resulted in a ~6gb .TDE file.
Since upgrading to 10.5 a few weeks ago, this same exact extract takes ~3 hours to create and results in a >10gb .HYPER file.
Is this the typical experience of others using 10.5, and are there any known specific scenarios in which .HYPER results in a longer extract time with a larger file whereas otherwise it would be roughly equivalent to .TDE? I'm not seeing a significant jump in performance, either, so I'm left scratching my head about what benefits there are that would prompt Tableau to implement this new architecture.
Hello Devin,
That is an unexpected behavior of the Tableau Desktop 10.5 .hyper feature. The extract should definitiely be at least somewhat faster but not slower.
Please open a case with the Tableau Technical Support teaml:
Support Case | Tableau Software
Please include as much detail as possible for the team to investigate. Including a copy of the original .tde file and .hyper file created would be fantastic.
Cheers,
Byrne, Patrick
