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 some workflows, we are heavily IO bound. If you watch the dask status page and see long read_${format} tasks, you might intuit that storage hardware is running slowly.
A nice way to expose storage performance to users could be via JupyterLab extension. I'm not sure if that would need to be part of this project or something separate, but GPUs starving for data is certainly something we need a way to observe.
The text was updated successfully, but these errors were encountered:
In some workflows, we are heavily IO bound. If you watch the dask status page and see long
read_${format}
tasks, you might intuit that storage hardware is running slowly.A nice way to expose storage performance to users could be via JupyterLab extension. I'm not sure if that would need to be part of this project or something separate, but GPUs starving for data is certainly something we need a way to observe.
The text was updated successfully, but these errors were encountered: