You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In post-processing, we often have less resources as in writing.
We often reach the ulimit -n for open files on Jupyter or ParaView or VisIt or scripts for sub-files.
This can often be work-arounded by increasing the ulimit -n ... soft-limit closer to the hard-limit of a machine.
Idea discussed today with @pnorbert@franzpoeschel:
Let's programmatically increase the soft-limit of a ADIOS2-enabled process to like 90% or 100% of the hard-limit for open files.
This has been added to master branch with #3376. BP3, BP4 and BP5 now will raise the softlimit to the hardlimit automatically.
Can you please test this out on your datasets. What about Windows? I made a blind call for windows to raise the limit but I don't know if it works.
In post-processing, we often have less resources as in writing.
We often reach the
ulimit -n
for open files on Jupyter or ParaView or VisIt or scripts for sub-files.This can often be work-arounded by increasing the
ulimit -n ...
soft-limit closer to the hard-limit of a machine.Idea discussed today with @pnorbert @franzpoeschel:
Let's programmatically increase the soft-limit of a ADIOS2-enabled process to like 90% or 100% of the hard-limit for open files.
X-ref:
The text was updated successfully, but these errors were encountered: