Ka-Hing Cheung

I think a naive solution probably works fine. So just a list of file names (separated by '\0') every time it's changed. We keep the list of dirty files in...

please include out of --debug_s3 and --debug_fuse

There was another PR https://github.com/kahing/goofys/pull/462 that did this. I've never used go mod at the time and sort of let it gather dust. I've since started using go mod for...

that line looks like seaweed is sometimes not returning content-length?

could you provide more info about what's lost and what the workload is like?

could you be more explicit? What operations did you do? What did you expect to see? What did you actually see?

Could you provide some sample code? At databricks we have many customers using tensorflow + goofys and it works fine. One caveat is that you cannot use hdf5 for model...

if you see this, then some read requests have failed. Depends on your application, it may or may not have retried. It doesn't seem like the workload should have triggered...

are you missing any permissions in azure? what does `--debug_s3` say?

any reason to use this instead of say, `--dir-mode=0777 --file-mode=0666`?