March 2006 Archives

This wasn't set correctly on geeklair.net (it was at some time, but I probably forgot to reset it when I recovered the machine).

On Mac OS X client, the only supported way to do this is via the GUI (since the Mac OS X Server CLI application one would use for this, systemsetup, isn't part of Mac OS X Client), which would be a problem for me since I don't have (and don't really want) remote GUI access set up for that machine.

Fortunately, the ARD stuff that comes with Mac OS X client now brings along a version (or two) of systemsetup.

Running: `cd /System/Library/CoreServices/RemoteManagement/ARDAgent.app/ && cd Contents/Support/ && ./systemsetup-panther -setrestartpowerfailure on`

Seemed to work. (systemsetup and system_profiler now both report that the machine will automatically restart on power loss).

Since I was stumped, I asked on darwinos-users, and got a reply from Shantonu Sen (who is smart and very active on the mailing lists).

I'm going to see if I can get a sample of 'securiytd' from when this happens next and that might point to something else that I might need to sample (lookupd, DirectoryServices ...)

Otherwise, I'll have to get the kernel debug kit installed and see if I can set up remote debugging and get it to work.

Fun fun.

Update:

Here's the sample from the stalled securityd -

Analysis of sampling pid 45 every 10.000000 milliseconds
Call graph:
    979 Thread_100f
      979 0x25dcc
        979 0x2ee4
          979 0x38d4
            979 0x8080
              979 0x83e8
                979 0x9054
                  979 0x9128
                    979 0x9acc
                      979 0x9c50
                        979 0xa094
                          979 0x3eac
                            979 pthread_mutex_lock
                              979 semaphore_wait_signal_trap
                                979 semaphore_wait_signal_trap
    979 Thread_1103
      979 _pthread_body
        979 0x16a48
          979 0x16ab8
            979 0x83e8
              979 0x9054
                979 0x9128
                  979 0x136b0
                    979 0x13868
                      979 0x13dac
                        979 0x19354
                          979 0x19354
    979 Thread_1203
      979 _pthread_body
        979 0x16a48
          979 0x16ab8
            979 0x83e8
              979 0x9054
                979 0x9128
                  979 0x9acc
                    979 0x9c50
                      979 0xa094
                        979 0x3eac
                          979 pthread_mutex_lock
                            979 semaphore_wait_signal_trap
                              979 semaphore_wait_signal_trap

Total number in stack (recursive counted multiple, when >=5):

Sort by top of stack, same collapsed (when >= 5):

        semaphore_wait_signal_trap        1958

        0x19354        979

And for no real reason at all, I'm starting to wonder if maybe the RAM on that machine is bad/going bad?

Lansing State Journal: Wikipedia world: Online reference site blurs fact/fiction line

I guess it's hard for people who work for a newspaper to understand something like Wikipedia.

(They should have noted how long it took for the extra crap the students added to be removed and also that you can look at all the different versions of the article and discussion about changes - something you can't do with a traditional encyclopedia).