I mounted an external drive as a shared resource and made it available on my local network using AFP. That machine and another machine use it for Time Machine backups. Eventually the machine where the drive is mounted starts acting strangely: almost no processes can be started. The console shows errors "kernel: file: table is full". Running lsof to see what's going on shows over 10,000 lines like this

AppleFile  5243            root *995u     IPv4 0xc4ee40a52bd9f69         0t0      TCP> (CLOSED)

I've already boosted the max files per process and system based on this answer and I don't think raising the limit further is the real solution. Is the idea of sharing a single drive between the two machines for Time Machine just a bad idea, or do you think this is possible to fix or a better configuration to be running. Both machines are on 10.9.5.

It wraps off the line above, but all the "bad" files are marked "CLOSED". Rebooting does not remove these files. Is there something I need to do to clear that up?


try loooking at port 61582 in the firewall / TM controller

Does that show as closed ? or at the very least does dmesg or your networking log(s) show anything to the same?

| improve this answer | |
  • Things got to the point that I decided to run diagnostics: too many things were going wrong and rebooting wouldn't clear anything up. I ended up physically power-cycling the machine (as in pull the cord) and there is no evidence of the problem I was having any longer. I'm starting to think that perhaps there's some hardware problems, but I've yet to run the diagnostics after all that. – Fitter Man Apr 14 '15 at 22:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.