All righty, let me bring things up to speed. I did a complete reinstall of Slack today (to the latest version), and backed up my config files before I did the reinstallation. So after the reinstall I have been getting as much as I could back up. I of course started with Samba, added the users, groups, permissions, and what not. And right now, everything appears to be working fine except 1 small problem. I can transfer everything to and from the machine without much problem. My mom can on hers except it transfer things A LOT slower. I even noticed a difference under top, where when I used Samba it took 15-20% of CPU power whereas mom's took .5-1%. Any ideas? Is it possible I overlooked something? The performance lag seems to only be noticeable in Samba...as I can ping around fine at good speeds, plus the cable is my internet lifeline and it operates fine. Any ideas?

Any ideas? Anyone have similar problems? I am running Samba version 3.0.20. I have noticed there is a newer version but haven't any luck upgrading to it and it isn't available on the mirrors yet.

Do the Samba (/var/log/samba) logs note anything while, or right before, it takes up so much of the CPU time?

These are the only 2 that strike my attention:


[2005/10/23 11:03:41, 0] printing/print_cups.c:cups_cache_reload(85)

[2005/10/23 11:01:20, 0] smbd/oplock.c:request_oplock_break(1081)
request_oplock_break: no response received to oplock break request to pid 1276 on port 32770 for dev = 303, inode = 3156, file_id = 218


I will say that this system does not use CUPS, although it's possible I may have installed it and let it run (oops on my part ). Not sure about the other error though.

Nevermind, fixed it.

And yup, it was something simple.

How's about posting the solution for people who are having similar issues?

My solution: check your NIC card settings to make sure they are all the same. Anytime I have ever used autosense I always run into problems such as draggy performance, collisions, etc.