-
Notifications
You must be signed in to change notification settings - Fork 3.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fileslower and filetop report TCP instead of file name #609
Comments
This is because it's tracing at the VFS layer, which catches file read/writes and other read/write types. This could either be filtered, or, documented. Perhaps the most intuitive option would be to filter by default, unless a verbose option is passed at the command line. |
I think this could be a consequence of using A few days ago I was investigating a writeback bug that slowed down I wrote a basic script to trace Here's the workload generator (write.py): import tempfile
import time
with tempfile.NamedTemporaryFile(prefix='/tmp/paddingpaddingpaddingpadding') as f:
while True:
f.write('test\n')
f.flush()
time.sleep(1)
The string in This doesn't work, though. I tried updating // smaller of the buffer or name length
int len = de->d_name.len < sizeof(val.name) ? de->d_name.len : sizeof(val.name);
bpf_probe_read(&val.name, len, (void *)de->d_name.name); but I get a huge error. Switching for @4ast, @drzaeus77 any suggestions for making this work? #188 looks a bit related but I'm not sure if it's the same underlying cause. |
This is similar in spirit to what was done in PR #677 to fix the problem reported in #609. filelife.py is now converted to use the right struct field. Signed-off-by: Marco Leogrande <[email protected]>
I believe this was fixed by recent commits. Closing ticket. |
I'm running it on 4.4.10.
The text was updated successfully, but these errors were encountered: