I'm quite convinced this is not an apparmor problem with interpreted
scripts. I created another test program that used wstatus and looped it
for many minutes and had no errors.
I took a look at the isc-dhcp source and while I can't be sure since I can't
reproduce this here, I think the bit of code that is triggering this is
bind_lease() from client/dhclient.c, specifically line 1292 that does:
1289 /* If the BOUND/RENEW code detects another machine using the
1290 offered address, it exits nonzero. We need to send a
1291 DHCPDECLINE and toss the lease. */
1292 if (script_go (client) == 2) {
script_go has:
3403 pid = fork ();
3404 if (pid < 0) {
...
3407 } else if (pid) {
...
3415 } else {
3416 /* We don't want to pass an open file descriptor for
3417 * dhclient.leases when executing dhclient-script.
3418 */
3419 if (leaseFile != NULL)
3420 fclose(leaseFile);
3421 execve (scriptName, argv, envp);
3422 log_error ("execve (%s, ...): %m", scriptName);
I still don't know what the problem, but I'm ok with adjusting the
access in the dhclient profile to allow reading of /bin/bash. I would
not be surprised if this didn't solve the connectivity issue.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1197484
Title:
Connection requests to saucy server VMs from a hosts fail after fresh
VM installs
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/isc-dhcp/+bug/1197484/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
