Freezes - EL Capitan or Security Spy
I am running 10.11.5 of El Capitan.
I have a Mac Mini Late 2012 with a SSD drive and 16gb of RAM. I have a clean install of El Capitan which was performed using a bootable OSX Apple Recovery disk which installed EL Capital directly rather than having to upgrade.
The only applications running on the machine are Security Spy and Plex Server.
After a lot of experimentation, full reinstalls and trial and error I have found the following.
If I set Security Spy to perform all motion detection and recording (which increases network traffic) - workload on Security Spy the OSX machine often runs for 1-2 days before it just freezes. By free I mean the network seems to stop. If I go to the console I can press a key to trigger the display to appear (power saving monitor off). The login screen is displayed but if you try and logon all you get is the beach ball and you have to do a full power off on to carry on.
If I set Security Spy to only view the cameras so that it's effectivy idling until someone uses a mobile app or web browser to look at a camera then the system remains stable for about a week before it happens again. When using this mode I switch the cameras to perform independent motion recording to a ftp server.
I have done a boot into hardware diagnostics and it did not find any fault. I have also looked at the system log but it does not show anything. Today for example the logs just stop at 2am. I did a reboot and it's all working now.
So - maybe the freeze is more frequent the more network i/o is occuring ?
I'm not convinced that this is a SS4 as it started before I upgraded. I have read about freezing issues with El Capitan 10.11.4 & 5 but I don't know if that's causing the issue or not as there is literally nothing to go on.
Today I signed up for the OSX beta and I've installed the latest 10.11.6 - so I will post an update as and when I have any news.
Does anyone else experience freezes like this ? I hope it's not a hardware issue as it's all out of warranty now and it would probably be cheaper to replace it that start a trial and error swap out of components.
Thanks in advance
Dave
I have a Mac Mini Late 2012 with a SSD drive and 16gb of RAM. I have a clean install of El Capitan which was performed using a bootable OSX Apple Recovery disk which installed EL Capital directly rather than having to upgrade.
The only applications running on the machine are Security Spy and Plex Server.
After a lot of experimentation, full reinstalls and trial and error I have found the following.
If I set Security Spy to perform all motion detection and recording (which increases network traffic) - workload on Security Spy the OSX machine often runs for 1-2 days before it just freezes. By free I mean the network seems to stop. If I go to the console I can press a key to trigger the display to appear (power saving monitor off). The login screen is displayed but if you try and logon all you get is the beach ball and you have to do a full power off on to carry on.
If I set Security Spy to only view the cameras so that it's effectivy idling until someone uses a mobile app or web browser to look at a camera then the system remains stable for about a week before it happens again. When using this mode I switch the cameras to perform independent motion recording to a ftp server.
I have done a boot into hardware diagnostics and it did not find any fault. I have also looked at the system log but it does not show anything. Today for example the logs just stop at 2am. I did a reboot and it's all working now.
So - maybe the freeze is more frequent the more network i/o is occuring ?
I'm not convinced that this is a SS4 as it started before I upgraded. I have read about freezing issues with El Capitan 10.11.4 & 5 but I don't know if that's causing the issue or not as there is literally nothing to go on.
Today I signed up for the OSX beta and I've installed the latest 10.11.6 - so I will post an update as and when I have any news.
Does anyone else experience freezes like this ? I hope it's not a hardware issue as it's all out of warranty now and it would probably be cheaper to replace it that start a trial and error swap out of components.
Thanks in advance
Dave
Comments
Mac-Mini mid-2011, 240gb SSD and 16 gb RAM. Mac OS 10.11.5.
Only two apps are always on: SS4 and PS3 Media Server (like Plex, but more simple )
There is 6 cameras:
- 2 cams continous record (1280x720)
- 2 cams with motion detection (armed all the time)
- 2 cams with motion detection (armed when I'm not at home)
Motion detection send an email when a motion is detected.
Video streams are recorded on a external USB disk (2To), not on the internal SSD disk.
Cams and mac-mini are connected on a gigabit switch (cams are connected to 100mbps but the mac is connected to 1gbps).
I see the latest Mac Mini is only dual core, not sure how a faster clock compensates for less cores.
04/07/2016 16:42:39.528 SecuritySpy[257]: AVF error: Perf capability check failed: 0xe00002be
These problems are the most difficult to track down, I don't envy you! There's nothing particularly about SecuritySpy that could trigger this, except that it uses system resources (RAM, disk, network etc.) so if there is a fault with your Mac it is likely to be shown more when something is running doing something.
If you installed the RAM recently, you could be looking at a bad chip. You could try taking the modules out a testing them one at a time. Is the SSD new too?
There is nothing in particular about OS X 10.11 that I know of that would make it susceptible to this, though you could try a different system version as a test (e.g. OS X 10.10).
We have also seen the "Perf capability check failed" message, and it has proven impossible to work out what it actually means! However as far as we can tell, it's a harmless message and can be safely ignored.
I thought it was a little warning like it checks for hardware video decoding and failed, but maybe not as your Mac I think should be able to do HW video decoding so must be something else...
A couple instances of terminal running yes command to thermal stress the CPU would verify the diagnosis in a few minutes. If you think it is a CPU thermal failure, open some terminal windows and run
yes > /dev/null
You'll see CPU usage skyrocket by about 1 entire core for each instance of yes. This lets you quickly see if you have a high temperature intermittent.
Be sure to close those terminal windows to kill the yes processes when done.
Anonymous UUID: 228ABA51-63DD-626B-AF64-685751F23768
Fri Aug 5 07:19:59 2016
*** Panic Report ***
panic(cpu 4 caller 0xffffff8022bce6fa): Kernel trap at 0xffffff7fa5305586, type 14=page fault, registers:
CR0: 0x000000008001003b, CR2: 0xffffff81e60b3004, CR3: 0x0000000025fe0000, CR4: 0x00000000001626e0
RAX: 0x0000000049453041, RBX: 0xffffff8048834600, RCX: 0x0000000000040004, RDX: 0x0000000000000003
RSP: 0xffffff820c6a3d28, RBP: 0xffffff820c6a3d30, RSI: 0x0000000000040004, RDI: 0xffffff80439f0000
R8: 0xffffff80491c8800, R9: 0xffffff804396cc00, R10: 0x00000000fffffffc, R11: 0xffffff81e6073000
R12: 0xffffff7fa5327a98, R13: 0xffffff80491c8800, R14: 0xffffff8043891300, R15: 0x0000000010001000
RFL: 0x0000000000010202, RIP: 0xffffff7fa5305586, CS: 0x0000000000000008, SS: 0x0000000000000000
Fault CR2: 0xffffff81e60b3004, Error code: 0x0000000000000002, Fault CPU: 0x4, PL: 0
Today I have replaced all the ram chips (2 x 8gb). Hopefully this will help. I know the units do run hot but I'm not sure if the thermal changes might cause the RAM chips to fail or maybe the socket / seating to expand and break the connection.
Fingers crossed.
Thanks for sharing your experiences here. I'm also running a Mac Mini Server with 16GB RAM. I'm running 6 cameras and am suffering a very similar situation to you. I'm starting to feel that the issue points towards El Capitan and SecuritySpy. I downgraded to Mavericks a month or so ago and everything seemed to work reliably. About 10 days ago, I stupidly thought I'd upgrade to El Capitan (I usually think it's best to run the latest OS whenever possible) but now I'm suffering from Mac freezes, the Mac crawling to a snails pace where mouse clicks take 10 secs + before menus open, apps quit etc. It's painful. I think it's now corrupted the Operating System as it now doesn't boot without it being painfully painfully slow. I also noted that SecuritySpy kept quitting and restarting regularly when it was running. All very very odd.
I've removed RAM chips, swapped them and it's made no difference. I still believe the hardware is fine (I used Apple's hardware test.)
Whilst this is a painful experience, I'm glad I'm not alone. I believe it's a SecuritySpy and El Capitan issue, but quite what I don't know.
There is plenty of discussion on the Apple forums about El Capitan issues... https://discussions.apple.com/thread/7250648?tstart=0
I'm about to downgrade the OS again. All very tiresome. Please let me know if you discover anything useful and I'll do similar.
Jonathan
All very strange.
Do you have an update Dave?
there are several pay-for ram testers with GUI, or the pretty cheap http://www.memtestosx.org/ that is command-line and can be run from single user mode (if you put it in the /bin/ directory) to test all but the first 200 MB or so of ram. Personally (likely from having been a tech for 20+ years) I run memtest on any ram I install in computers, or on other peoples computers that are having weird issues. It may only be one in every 50 to 100 that is ram at fault, but if the ram isn't verified as being good it could be causing all kinds of weird issues. (I also usually run memtest overnight or a full day when possible doing several many tests
I do run SMC Fan Control to keep my fans running a bit higher than normal to keep the temp stable on my Mac mini 2012 (2.3 Ghz i7) - with fully default (no SMC fan control) the temperature fluctuates quite a bit, and the fans would go from idle speed to quite fast every so often (sometimes every few minutes, sometimes might be 15-20 minutes for the cycle) - in SMC Fan Control I set a minimum fan speed of 3300 rpm and most of the time it stays steady at that speed - it can still run the fans faster if needed.
With 10.10 I found I had to reboot every couple of weeks or eventually the system would crash because of the video drivers - with 10.11 (El Capitan) the integrated intel video drivers are more stable I've frequently gone beyond a month and am usually restarting for other things like software updates, maybe only once or twice have seen video artifact issues. I did wait to go to El Capitan until 10.11.2 or .3 there are typically issues with releases for .0 and .1 (my Mac mini runs security spy, OS X Server for Caching server & web server, Plex - including frequent transcoding of video, as well as more typical apps)
Hopefully something here can help you guys find the issue.
Replaced RAM
Used external USB Fan (which reduced the ambient internal temp by 10'c)
I have also updated to the latest EL Capitan.
Anonymous UUID: 228ABA51-63DD-626B-AF64-685751F23768
Fri Aug 19 22:22:19 2016
*** Panic Report ***
panic(cpu 0 caller 0xffffff80081ce6fa): Kernel trap at 0xffffff80082f6938, type 13=general protection, registers:
CR0: 0x000000008001003b, CR2: 0x000000010f1e5000, CR3: 0x0000000009ce6001, CR4: 0x00000000001626e0
RAX: 0xffffff80418e8c50, RBX: 0xffffff80272fda08, RCX: 0x7fffff80418e8b40, RDX: 0xffffff80272fda58
RSP: 0xffffff81f174bd30, RBP: 0xffffff81f174bdd0, RSI: 0x000000000000e000, RDI: 0xffffff80272fda20
R8: 0x0000000000000001, R9: 0x0000000000000000, R10: 0x0000000000002000, R11: 0x000000000000000a
R12: 0xffffff80418e8c30, R13: 0x0000000000000000, R14: 0xffffff80272fda20, R15: 0xffffff80272fda58
RFL: 0x0000000000010202, RIP: 0xffffff80082f6938, CS: 0x0000000000000008, SS: 0x0000000000000010
Fault CR2: 0x000000010f1e5000, Error code: 0x0000000000000000, Fault CPU: 0x0, PL: 0
mp = 0xffffff80272fda08, phys = 0x9423a08, prev (0x7: 0x7f13000-0x8000000)
vp = 0xffffff80418e8c30, phys = 0x13c12bc30, prev (0xb: 0xffe70000-0xffea0000)
0xffffff80272fc000: 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
0xffffff80272fd000: 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
0xffffff80272fe000: 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
Backtrace (CPU 0), Frame : Return Address
0xffffff81c6fd7c50 : 0xffffff80080dab52
0xffffff81c6fd7cd0 : 0xffffff80081ce6fa
0xffffff81c6fd7eb0 : 0xffffff80081ec563
0xffffff81c6fd7ed0 : 0xffffff80082f6938
0xffffff81f174bdd0 : 0xffffff8008527f73
0xffffff81f174bec0 : 0xffffff800830feed
0xffffff81f174bee0 : 0xffffff80082fa0b4
0xffffff81f174bf50 : 0xffffff8008304ea7
0xffffff81f174bf60 : 0xffffff8008628701
0xffffff81f174bfb0 : 0xffffff80081ecd66
BSD process name corresponding to current thread: launchd
Dave, have you tried downgrading to Mavericks?
Jonathan
7:50 up 32 days, 14:22, 3 users, load averages: 2.47 2.69 2.58
Mac OS X 10.11.6 (15G31), Security Spy 4.0.4 then 4.0.5 the last day or so.
For DaveGarret - it shows that LaunchD was the process running on that CPU core when it panic'd.
In case there is a pattern I have a upgraded SSD drive fitted and 16gb of ram.
Dave.
completely unrelated to SecuritySpy
Check for firmware update for your SSD (backup data/clone drive before doing a firmware update on SSD if there is one, just to be safe)
As a tech if something isn't stable, I start with ram test (which also tests CPU to some extent) if that passes at least an overnight run of repeated tests, then I test the hard drive next (no free software to do this unfortunately), if that passes as well, I run the system off of an external drive attempting to reproduce the issue - if it doesn't, it could be a software issue with the particular install of the OS, try a backup format & re-install of the OS - if it does reproduce then its harder, could be something wrong with the system that isn't showing up in the ram/hard drive tests - flakey cable/drive/power supply.
(My mothers 2012 Mac mini did have an issue last year with some freezes, and eventually more and more frequent sudden power offs usually when she wasn't using it, eventually it wouldn't turn on - applecare warranty replacement of power supply and it was back to normal operation, would have gotten it fixed sooner, but she didn't tell me about most of the problems until it wouldn't turn on any more)