Hi ,
We are recently migrated our Tableau server from Windows to Linux OS
Tableau server environment (Azure VM)
Environment- Tableau Server on Ubuntu 22.04 LTS
- 96 vCPU, 672 GB RAM
- 2‑node cluster
- Node 1: Primary + File Store
- Node 2: Backgrounder only (no File Store)
- Data disk mount (ext4) - 16 TB SSD premium P80 (900 MB IOPS)
After migration, extract refresh performance has degraded significantly:
- On Windows, the extract refresh used to take 45–60 minutes.
- On Linux, the same extract now takes ~120 minutes.
- From the logs, copying the .hyper file from the extract folder to the temp directory alone takes ~90 minutes.
- When a backgrounder job runs on Node 2, the copy of the extract from Node 1 → Node 2 is extremely slow, causing the entire refresh to take much longer.
***As checked swap 0 ***
#Tableau Server
8 respuestas
Hi @Emeric LE SAULNIER DE SAINT JOUAN
, Apologies for the delayed response
The time taken during copy hyper file from /extract folder to /temp folder.
After copy to Temp folder, the query execution, ingest data in hyper and write back to extract folder is quite fast/
temp & data engine folder on same mounted disk .
