Hi all,
We're running Tableau Server on Linux and recently upgraded to version 2024.2.7 from 2023.1.8. After the upgrade, I noticed that the disk usage on one of our volumes has jumped significantly — over 880 GB being consumed in /var/opt/tableau/tableau_server/data/tabsvc/dataengine/extract/
.
To dig into it, I started inspecting the extract folders and noticed something strange. For example, we have a data source called "Sample View With Active Allocation" (name anonymized), and under the TSM there are three different folders, each containing a .hyper file with the exact same name
Each file is 27 GB, and they live under different UID paths, like:
- {8BB5...}
- {9F56...}
- {BE84...}
These appear to be duplicates — possibly leftover from previous extract refreshes — but I only see one published extract in the Tableau UI, which is the current one scheduled for refresh. I’ve observed this pattern with multiple other extracts as well — many of them seem to have several large .hyper files lingering, all with the same name, and each consuming tens of gigabytes.
This behavior did not happen in our previous version (2023.1.8). It seems to be new in 2024.2.7.
My questions are:
- Which Tableau Server component or process is responsible for retaining these extra .hyper files?
- Could this be a new bug or expected behavior in 2024.2.7?
Thanks in advance for your help!
@Mudassar Shareef
Hi, apply the first option:
tsm restart
Also, I suggest you to update your Tableau Server to a newer subrelease.
@Emeric LE SAULNIER DE SAINT JOUAN do you have more comments on this?
If this post resolves the question, would you be so kind to "Select as Best"?. This will help other users find the same answer/resolution and help community keep track of answered questions. Thank you.
Regards,
Diego Martinez
Tableau Visionary and Forums Ambassador
