Improving Performance + Dataset Chunk Cache Management - Aleksandar Jelenak - Jan 14 Call the Doctor

Improving Performance + Dataset Chunk Cache Management - Aleksandar Jelenak - Jan 14 Call the Doctor

In our next Call the Doctor session on Tuesday, January 14, The HDF Group’s resident scientist for the scientists, Senior Informatics Architect Aleksandar Jelenak (@ajelenak ) will talk about some ways to improve performance and dataset chunk cache management.

  • What library features could improve performance when having a lot of groups in a file? Is this a good practice?
  • Dataset chunk cache management.
    • When is a chunk cache created?
    • When is a chunk cache destroyed?
    • What happens with the memory occupied by a chunk cache when it is destroyed?
    • Is there any scenario when the library would somehow keep the memory of destroyed chunk caches for later reuse?

To join, just jump on the zoom:
Launch Meeting - Zoom
January 14, 2024,12:20 p.m. central time US/Canada

:arrow_right: Reminder: The Working Group series meeting is currently suspended. No meeting January 9th.