how long do you need to shut down your system?

[dyn]

Dyn, With a 2012 SSD MBP, bootup times are about 8 seconds with any OS X version. I have never seen a spinning-HD OS-X system boot up nearly this fast - more like 30 sec. - 1 min. Mine takes 5-8 seconds up, and 3-5 seconds down. It will restart from running back to useable config in less than 15 seconds.

Those are nice boot times and shutdown times indeed but it's not what I meant with that. I've seen quite a number of computers with Windows, Linux and OS X that have a boot up time that requires grabbing coffee. In most cases booting up takes longer than shutting down and it's just odd to see people not compain about something that takes longer than the one they are complaining about. That's intriguing. What makes shutdown so important and boot up so unimportant?

You raise a good point; if an SSD OS X machine started taking 30+ seconds to boot, I believe people would definitely raise hell about it. I certainly would.

That's what I thought but from an experience I can tell you that doesn't really happen in the real world. Most seem to be complaining when it takes more than getting some coffee or going to the loo (thus more than 5 minutes) for both boot up and shutdown. I would complain too if it would take more than 30 seconds (if there are no updates to be installed). I've even come across quite a few users who were perfectly fine with very slow machines (it takes them 15 to 30 minutes to go from on to being logged in...yes, I pondered sending them to a psych..).
From a user point of view I can understand it a little bit. How long should a computer boot up/shutdown? Is it 10 seconds, 30, 90? They have no clue since they have no clue about IT. That might explain why they are not complaining. They just don't know it should/can be faster.

I agree with KB - Shutdown monitor repeatedly tries to kill the launchd and launchagents processes that are notorious for hanging on shutdown. Some of these processes previously claimed not to support "dirty/idle" exits. One of these was launchd.peruser.501 that also did not respond to a 1 sec. terminal hack timeout with ML 10.8.4-5 and Mav. DP4, when every other problematic process was successfully remedied by 1 sec. terminal timeout hacks. Therefore, it is highly noteworthy that Apple has remedied these processes to have a successful default 1 second kill with 10.9 DP5.

I wonder how much 3rd party apps also cause this issue. Not every app supports something like sudden termination which Apple has implemented since 10.7 and uses more strictly in 10.8. Could be that this might lead to problems as well. Since 10.9 again changes a lot on how to handle apps running in the background and such it could be that Apple tweaked and we get better shutdown times because of that.

 

[m4v3r1ck]

With OS X 10.8.4 (12E55) on my Mac Pro 3,1 it seems that my back-up service (KPN) is slowing down or even preventing the shut down process! Wicked...

 

[ATC]

What would you define as being "lengthy"? I'd say that anything within 15 seconds is fast, within 30 seconds is normal (until complete power down that is). These shutdown times are seen with other desktop/laptop systems (non-OSX ones). Obviously shutdown times depend on your settings as well as the software that was running when you hit shutdown. If it has to go through a lot of them it will take longer then when none was running.

I guess it's all relative. Based on your post then mine (non-SSD, 7200rpm HD) would be on the slow side of normal. I'm getting 20 second shutdown from a clean desktop (nothing running) or log in screen. I recall the early builds of ML (10.8, 10.8.1 and maybe .2) I was getting under 10 second shutdowns. If I have only the 4 core Apple apps running (Mail, Calendar, Safari and Messages) shutdown is between 30 and 40 seconds. No big deal but whatever Apple did since 10.8.4 (maybe since 10.8.3, I just can't recall exactly when it started happening) and now 10.9 my shutdown times have gone up over 100%.

To be fair, I really shouldn't be complaining about shutdown times for two reasons; one, I don't shut down often enough to really notice, I use sleep almost always. Second, my boot times are way worse, in the 50 second territory. So if anything I should be complaining about boot instead of shutdown times.

 

[Krazy Bill]

I wonder how much 3rd party apps also cause this issue.

None. To confirm this, shutdown from your login screen before anything is even loaded. It still takes the same amount of time.

 

[ultraspiracle]

None. To confirm this, shutdown from your login screen before anything is even loaded. It still takes the same amount of time.

Very good reply, that.....

 

[Vinz-uk]

Slow Shutdown

Hi Everyone.

I also had an issue with a slow shutdown from ML 10.8.2.
Before that, the mac shut down very quickly.

Running the verbose mode, I saw that the shut down hang about 20s doing nothing with a process called "appleeventsd".

The solution found, looking on different forums si to launch 3 commands, one after each other in the terminal to kill this process after 2 seconds:


sudo launchctl unload /System/Library/LaunchDaemons/com.apple.coreservices.appleevents.plist

sudo defaults write /System/Library/LaunchDaemons/com.apple.coreservices.appleevents ExitTimeOut -int 2

sudo launchctl load /System/Library/LaunchDaemons/com.apple.coreservices.appleevents.plist

And then, your ML or Mavericks Mac should shut down very quickly again

 

[Krazy Bill]

The solution found, looking on different forums si to launch 3 commands, one after each other in the terminal

Yes. This hack has been floating around for a while but since Mavericks is in preview stages it's best to keep the hacks out to see if Apple fixes it on their own (which they have to some degree).

to kill this process after 2 seconds:

Why not 1 second?

 

[mmomega]

My shut down times .... I seriously could not say because I can not remember clicking the shutdown command on any of my Macs in quite sometime.

I have some Mac Mini's at work that have not been turned off in 3 years, not even power outages, they are on double battery backups.

 

[dyn]

None. To confirm this, shutdown from your login screen before anything is even loaded. It still takes the same amount of time.

Provided that no 3rd party tool has installed any launchagents and such that launch at boot time (not at login). I don't think this will be the case for many of us though.

Why not 1 second?

I'm not a big fan of using any hacks without knowing what it does. Doesn't have to be very detailed, just basics is enough. So yeah, I'm wondering the same thing: why 2 seconds and not 1 or 15ms or something else?

 

[Krazy Bill]

Provided that no 3rd party tool has installed any launchagents and such that launch at boot time (not at login). I don't think this will be the case for many of us though.

Do a shutdown from the login screen after a clean install. Same results.

This anomaly is *all* Apple. Always has been.

 

[Vinz-uk]

I have used this "hack" on my old MBP 2010 and on my rMBP for more than 6 months on Mountain Lion, and I have had no issues at all since then.
So I'm doing the same on Mavericks since Apple doesn't do anything to improve the shut down time
On my "hack", I changed the 2 sec time by 1sec, and it's event faster

 

[dyn]

Do a shutdown from the login screen after a clean install. Same results.

That is what I would advise too but as you can see, there are much more variables that need to be excluded before you can tell what's going on and what we can do to fix it.

This anomaly is *all* Apple. Always has been.

Problem is that only a certain group of users have this problem, most do not. There is something strange going on and Apple needs to address this. However let's not judge too quickly. I've seen people do that here too often. There have been cases where non-Apple stuff was the culprit of a slow shutdown or where the machine simply was slow in every facet. Something else was the problem there. That is why it is absolutely necessary to do some scientific tests.

 

[Krazy Bill]

but as you can see, there are much more variables that need to be excluded before you can tell what's going on and what we can do to fix it.

What other variables can a user take blame for after a clean install? (Take your time - I'll wait here).

Problem is that only a certain group of users have this problem, most do not.

Reproducible and documented timeout calls from EVERY mac out of the box since 10.8.2? My friend, this has been going on for some time. I suggest you do your homework.

That is why it is absolutely necessary to do some scientific tests.

Nothing scientific about checking your console logs. Dude, no offense but you are dismissed.

 

[SlCKB0Y]

You need to ask yourself if a shutdown in 2~3 seconds is the correct way of shutting down. When you run services, applications, etc. and you hit shutdown these need to be gracefully shutdown. If you don't you'll end up with data loss, data corruption...you know, nasty things.

Your whole post is one big fail since OS X pre Mountain Lion used to shutdown in less than 5 seconds (at least on SSD systems) and there are hundreds of blogs, webpages and forum posts complaining about this bug.

 

[Mike MA]

Yes, the 10.8.4 and 10.8.5 releases all suffer from slow shutdown.

It started with 10.8.1. Just for the record.

 

[dyn]

What other variables can a user take blame for after a clean install? (Take your time - I'll wait here).

Many. If you want examples simply browse the forums here. But that wasn't my point. I was pointing at the technical part of applications and operating systems. The part where no user or sysadmin can do anything about. If you have crappy software and you install it... That's all it takes.

Reproducible and documented timeout calls from EVERY mac out of the box since 10.8.2? My friend, this has been going on for some time. I suggest you do your homework.

I've clearly done mine, you clearly haven't done yours. This problem is NOT reproducible nor documented for every Mac out of the box since 10.8.2. This problem has been reproducible one some Macs every since 10.7 so yes it has been going on for quite some time.

Nothing scientific about checking your console logs. Dude, no offense but you are dismissed.

There's also nothing scientific about your posts. The only thing you do is trying to shout from the top of your lungs. That doesn't help anybody with this problem. I suggest you take these people serious. This is an annoying problem for some which needs to get resolved. In order to do that we need to gather some data and see where things are going wrong. This also means that you need to understand the data. Not everybody with a shutdown of 30 seconds has this problem. That way of doing things is what's scientific.

Your whole post is one big fail since OS X pre Mountain Lion used to shutdown in less than 5 seconds (at least on SSD systems) and there are hundreds of blogs, webpages and forum posts complaining about this bug.

No it didn't do that before Mountain Lion. Lion also has these shutdown issues. Also, quantity doesn't tell you much. It's what's inside those messages. With each new update or upgrade people have problems that they relate to the update/upgrade. In about 90+ % of the time they drew the wrong conclusion (it was just a coincidence the problem arose after the update/upgrade, there is no relation between the two). That's why you always have to check first and only then draw a conclusion, not the other way around like you are doing.

 

[Krazy Bill]

If you want...

Dude. You're finished. Done. Dismissed.

You're defending something for which you haven't got the technical skills to defend.

 

[ultraspiracle]

On my system, Lion had the same 20sec timeout kills programmed in, but actually shut down in less than 5 sec. The problem arose in ML 10.8.2 and wasn't fixed until 10.8.5. Even the first releases of Mav's 10.9 suffered from it. Apple themselves then SET THE DEFAULT TIMEOUTS to 1 sec and, presumably along with some other unknown adjustments, fixed the problem. This is confirmed by looking at shutdown logs as KB stated above.

That being said, according to my own experience that if someone is still experiencing slow shutdowns with Mav's 10.9.1 it has to be because of some 3rd party software conflict or some weird user account-specific issue.

Many. If you want examples simply browse the forums here. But that wasn't my point. I was pointing at the technical part of applications and operating systems. The part where no user or sysadmin can do anything about. If you have crappy software and you install it... That's all it takes.


I've clearly done mine, you clearly haven't done yours. This problem is NOT reproducible nor documented for every Mac out of the box since 10.8.2. This problem has been reproducible one some Macs every since 10.7 so yes it has been going on for quite some time.


There's also nothing scientific about your posts. The only thing you do is trying to shout from the top of your lungs. That doesn't help anybody with this problem. I suggest you take these people serious. This is an annoying problem for some which needs to get resolved. In order to do that we need to gather some data and see where things are going wrong. This also means that you need to understand the data. Not everybody with a shutdown of 30 seconds has this problem. That way of doing things is what's scientific.


No it didn't do that before Mountain Lion. Lion also has these shutdown issues. Also, quantity doesn't tell you much. It's what's inside those messages. With each new update or upgrade people have problems that they relate to the update/upgrade. In about 90+ % of the time they drew the wrong conclusion (it was just a coincidence the problem arose after the update/upgrade, there is no relation between the two). That's why you always have to check first and only then draw a conclusion, not the other way around like you are doing.

 

[MacMan988]

I usually shut it down as soon as my work is over on it.