Seeing as you didn’t ask, allow me to give you a tip: if you are ever in a hospital A&E department, never comment that it is quiet. I have known staff almost get lynched for that, which invariably results in half a dozen ambulances arriving with patients at death’s door.
Similarly, when you write or talk about computers, never articulate that you think the problem is solved. It won’t be, by virtue of your stating that.
You will recall that a couple of nights ago, my iMac apparently restarted in the dead of night. I investigated, couldn’t find anything obviously wrong, was not particularly helped by the logs, and retired with an uninterpretable
Previous shutdown cause: -128 message. Well, it happened again today, right in front of my eyes.
One minute I was happily working on an article for this blog, which will now have to be held over a little, listening to an audiobook in iTunes (The Epic of Gilgamesh, seeing as you didn’t ask that either). Then everything froze completely: no pointer, no keyboard, no clock ticking away the seconds. Before I could press and hold the Power button to force a shutdown, the display went black and my iMac restarted of its own accord.
I went straight to the logs, using Console, and could not find anything helpful immediately prior to the freeze. It looked like a very hard crash, so hard that the kernel couldn’t even say goodbye with a panic, and when restarted there was that mysterious
Previous shutdown cause: -128 again.
Given that this is a recent problem, I browsed the logs, and was surprised to see daily
awdd reports, usually fairly scrambled, since 5 March. But checking the InstallHistory, neither that, nor this pair of restarts, seemed to correlate with anything. I wasn’t getting any closer.
As this iMac had quite a few older things inherited by migration from my previous iMac (and it, in turn, from my Mac Pro 8-core), I did a bit of pruning of old cruft from various Library folders, and disabled or turned off anything which looked unneeded and wobbly. This is a dangerous process, and sure enough what I had done prevented my iMac from restarting. Not only that, but when I forced it to shut down, the external Thunderbolt Promise RAID remained powered up, and would not shut down either.
Things were not looking good: my Mac would not start up, and the RAID would not shut down. Stalemate.
I disconnected everything and, with a wired USB keyboard at the ready, went through my dead Mac routine. I seem to be getting pretty slick at doing it now, it must be all the practice I am getting. I first reset its SMC and NVRAM, then started it up in Apple Diagnostics – for which a wired USB keyboard is essential. Thankfully that could not find a hardware fault, so I then restarted in Recovery Mode, ran Disk Utility, and checked the Fusion Drive.
The only way that I could kill the Promise RAID drive was to pull its mains plug – not something that I like to do, but it is supposed to be designed to cope with such things.
Gradually I put everything back together again, and seem to be back to where I was before the automatic restart. It is tempting to blame something which blew in with the El Capitan 10.11.4 update, or the inevitable iTunes bug, but I don’t have any good evidence to support either claim.
The one slight worry that I have is that one event which does correlate with the start of these daily
awdd logs – which indicate a fairly hefty crash even if it does not require a restart – is the time that I set my Energy Saver preferences using MacPilot, to stop my Mac from sleeping its hard drive. I have reverted to the Energy Saver pane controls to see if they make any difference. Perhaps the reason that Energy Saver has insisted on sleeping my hard drive is that this model doesn’t like system sleep with the hard drive still spinning. But that shouldn’t have caused such a hard crash when nothing was in danger of going to sleep.
No doubt there will be further developments in the coming days…