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
Recently, we have deployed dynolog in our gpu cluster to collect trace files via kineto on-demand profiling. It needs extra efforts to collect trace files dumped to local storage via kineto for distributed applications. We saw that kineto supports dumping traces files to remote storage in https://github.com/facebookincubator/dynolog/blob/main/docs/pytorch_profiler.md, which is exactly what we want. But there's no other docs or tutorials introduce how to use remote storage. Could you provide an introduction or a tip on how to configure kineto to write trace files to remote storage?
The text was updated successfully, but these errors were encountered:
Hi, sorry I missed this earlier. The easiest way to dump files to a remote storage would be to create a logger that inherits ChromeTraceLogger and wraps up the parent finalizeTrace with a function that also uploads it to the desired remote path
Recently, we have deployed dynolog in our gpu cluster to collect trace files via kineto on-demand profiling. It needs extra efforts to collect trace files dumped to local storage via
kineto
for distributed applications. We saw that kineto supports dumping traces files to remote storage in https://github.com/facebookincubator/dynolog/blob/main/docs/pytorch_profiler.md, which is exactly what we want. But there's no other docs or tutorials introduce how to use remote storage. Could you provide an introduction or a tip on how to configure kineto to write trace files to remote storage?The text was updated successfully, but these errors were encountered: