12.10 64-bit - Nautilus spins and spins on Home, sometimes locks machine up

[ewangr]

I've been trying to diagnose this for the last day, and can't seem to figure out what is going on. When the machine first comes up everything is fine. After a period of time, when I go to open my home directory (i.e. /home/user), Nautilus just spins and spins. If I navigate away or click another directory it's fine. If I wait too long, the entire machine locks up. Nothing in syslog seems obvious, which implies that when it locks up the system isn't writing anything before I'm forced to power off.

Any ideas?

System hardware is a core i7 with 6 gigs of RAM, 128 gig SSD, wireless network adapter (works fine ever since i disabled ipv6). Attached externally is a 4 disk RAID array that automounts through /etc/fstab.

Thanks in advance.

[ewangr]

Just happened again, and again SysLog just shows the gap - no obvious errors. Any ideas of even where to look would be appreciated. I'm suspecting network - but that's just a guess based on where I've seen other comments.

Also, any reason to think ringtail would be more stable at this point?

[ManamiVixen]

What do you have in the folder? Nautilus starts to choke if there are more than 1,500 files in a single folder. The limit isn't recursive so it's not total, just per folder. For me at least that happens.

[ewangr]

At that level I just have the default folders, the usual hidden files, and the mount point for my Share. Well under 100 files. Again, it works just fine when it first comes up, it's only later (an hour or more) that "something" happens. Once that does if I can quickly close out, and remember not to open Nautilus using the default launcher, I seem to be fine. Although once it happens if I try to do ls -al in a terminal session for the home directory that will clock as well.

[rnerwein]

I've been trying to diagnose this for the last day, and can't seem to figure out what is going on. When the machine first comes up everything is fine. After a period of time, when I go to open my home directory (i.e. /home/user), Nautilus just spins and spins. If I navigate away or click another directory it's fine. If I wait too long, the entire machine locks up. Nothing in syslog seems obvious, which implies that when it locks up the system isn't writing anything before I'm forced to power off.

Any ideas?

System hardware is a core i7 with 6 gigs of RAM, 128 gig SSD, wireless network adapter (works fine ever since i disabled ipv6). Attached externally is a 4 disk RAID array that automounts through /etc/fstab.

Thanks in advance.

hi
my first question is: is the mountpoint of your RAID in your home ?
if nautilus spins then open a terminal.
follow these commands:
ps -ef | grep nautilus
sudo strace -p PID_OF_NAUTILUS (to quit the trace hit Ctrl c)
in the trace you will see something like this: read(3, 0x7fff590ecef0, 16) if you see always the same command in the trace then that is the point of loop !
where 3 is the file descriptor. then you can:
cd /proc/PID_OF_NAUTILUS/fd
ls -l (you can see e.g. where the fs 3 points to - hope this give you a hint).
hope it helps you
ciao

[tgalati4]

I would remove the external array and see if that improves the behavior. It could be an interaction between the external array and the system. Also, check the condition of your power supply. If the power supply drops out, your disk will slow down causing IOWaits, but not necessarily syslog entries. Use top to see if IOWaits are greater than 10% when the system exhibits the spinning ball of death.

[ewangr]

hi
my first question is: is the mountpoint of your RAID in your home ?

The raid mountpoint is in /media/ewang, so while it could be an issue, I don't think that's it. The share drive (downstairs computer) is mounting on my home directory though, so perhaps that's an issue?

if nautilus spins then open a terminal.
follow these commands:
ps -ef | grep nautilus

results:
ewang 1964 1845 0 18:20 pts/1 00:00:00 grep --color=auto nautilus
ewang 3420 1 0 Mar24 ? 00:00:53 nautilus

sudo strace -p PID_OF_NAUTILUS (to quit the trace hit Ctrl c)

results:
recvfrom(6, "U\2\304.\323?{\4\3\4\4\0\0\0\0\0\0\0\0\4\4\4\4\4\ 0\0\3\37%\2\0\0", 4096, 0, NULL, NULL) = 32
recvfrom(6, 0xd3a8b4, 4096, 0, 0, 0) = -1 EAGAIN (Resource temporarily unavailable)
recvfrom(6, 0xd3a8b4, 4096, 0, 0, 0) = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=5, events=POLLIN}, {fd=6, events=POLLIN}, {fd=13, events=0}, {fd=3, events=POLLIN}, {fd=12, events=POLLIN}, {fd=9, events=POLLIN}], 6, -1) = 1 ([{fd=6, revents=POLLIN}])
recvfrom(6, "U\2\304.\313B{\4\3\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\ 0\0\3\37%\3\0\0", 4096, 0, NULL, NULL) = 32
recvfrom(6, 0xd3a8b4, 4096, 0, 0, 0) = -1 EAGAIN (Resource temporarily unavailable)
recvfrom(6, 0xd3a8b4, 4096, 0, 0, 0) = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=5, events=POLLIN}, {fd=6, events=POLLIN}, {fd=13, events=0}, {fd=3, events=POLLIN}, {fd=12, events=POLLIN}, {fd=9, events=POLLIN}], 6, -1) = 1 ([{fd=6, revents=POLLIN}])
recvfrom(6, "U\2\304.\333F{\4\3\4\4\0\0\0\0\0\0\0\0\4\4\4\4\4\ 0\0\3\37%\2\0\0", 4096, 0, NULL, NULL) = 32
recvfrom(6, 0xd3a8b4, 4096, 0, 0, 0) = -1 EAGAIN (Resource temporarily unavailable)
recvfrom(6, 0xd3a8b4, 4096, 0, 0, 0) = -1 EAGAIN (Resource temporarily unavailable)
poll([{fd=5, events=POLLIN}, {fd=6, events=POLLIN}, {fd=13, events=0}, {fd=3, events=POLLIN}, {fd=12, events=POLLIN}, {fd=9, events=POLLIN}], 6, -1^C <unfinished ...>

in the trace you will see something like this: read(3, 0x7fff590ecef0, 16) if you see always the same command in the trace then that is the point of loop !
where 3 is the file descriptor. then you can:
cd /proc/PID_OF_NAUTILUS/fd
ls -l (you can see e.g. where the fs 3 points to - hope this give you a hint).

results:
lr-x------ 1 ewang ewang 64 Mar 24 21:34 0 -> /dev/null
l-wx------ 1 ewang ewang 64 Mar 24 21:34 1 -> /dev/null
lrwx------ 1 ewang ewang 64 Mar 24 21:34 10 -> anon_inode:[eventfd]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 11 -> anon_inode:[eventfd]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 12 -> socket:[20764]
lr-x------ 1 ewang ewang 64 Mar 24 21:38 13 -> /proc/3420/mounts
lr-x------ 1 ewang ewang 64 Mar 24 21:38 14 -> /home/ewang/.local/share/gvfs-metadata/home
lr-x------ 1 ewang ewang 64 Mar 24 21:34 15 -> /home/ewang/.local/share/gvfs-metadata/home-581d3f80.log
lrwx------ 1 ewang ewang 64 Mar 24 21:34 16 -> socket:[17878]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 17 -> anon_inode:[eventfd]
lr-x------ 1 ewang ewang 64 Mar 24 21:34 18 -> /proc/3420/mountinfo
lrwx------ 1 ewang ewang 64 Mar 24 21:34 19 -> socket:[16834]
l-wx------ 1 ewang ewang 64 Mar 24 21:34 2 -> /home/ewang/.xsession-errors
lr-x------ 1 ewang ewang 64 Mar 24 21:34 20 -> /home/ewang/.local/share/gvfs-metadata/root
lrwx------ 1 ewang ewang 64 Mar 24 21:34 21 -> anon_inode:[eventfd]
lr-x------ 1 ewang ewang 64 Mar 24 21:34 22 -> /home/ewang/.local/share/gvfs-metadata/root-801fc16b.log
lr-x------ 1 ewang ewang 64 Mar 24 21:34 23 -> /home/ewang/.local/share/gvfs-metadata/uuid-c0e21aeb-28b1-4b28-9bff-071e78a2f6c4.AW14TW (deleted)
lr-x------ 1 ewang ewang 64 Mar 24 21:34 24 -> /home/ewang/.local/share/gvfs-metadata/uuid-c0e21aeb-28b1-4b28-9bff-071e78a2f6c4-ee59ee3f.log (deleted)
lrwx------ 1 ewang ewang 64 Mar 24 21:34 25 -> socket:[14075]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 26 -> anon_inode:[eventfd]
lr-x------ 1 ewang ewang 64 Mar 24 21:34 27 -> /home/ewang
lrwx------ 1 ewang ewang 64 Mar 24 21:34 3 -> socket:[14071]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 4 -> anon_inode:[eventfd]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 5 -> anon_inode:[eventfd]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 6 -> socket:[14072]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 7 -> socket:[16324]
lrwx------ 1 ewang ewang 64 Mar 24 21:34 8 -> anon_inode:[eventfd]
lr-x------ 1 ewang ewang 64 Mar 24 21:34 9 -> anon_inode:inotify

Let me know if any of the above points to something obvious.

Thanks!

[tgalati4]

Was there anything helpful in these logs?

/home/ewang/.local/share/gvfs-metadata/home-581d3f80.log
/home/ewang/.local/share/gvfs-metadata/uuid-c0e21aeb-28b1-4b28-9bff-071e78a2f6c4-ee59ee3f.log (deleted)

It's possible that if you have a slammed network (i.e. Neflix streaming) then your network share could be causing the hang up. Try removing the share for a couple of days and see the behavior changes.

[Singtoh]

hello ewangr,

I am using ubuntu 12.10 64 with gnome shell. I am just making an observation here but the same thing happens on my machine. The way I stop this from happening is go to advanced settings(gnome-tweak-tool) Desktop, and in that menu toggle off "Have file manager handle the desktop" and the spinning cursor dissapears and everything is normal again. I just turn it back on when I want to change wallpaper then turn it off again. Maybe your problem is related to the file manager handling the desktop??? Not sure, just thought I would put in my two cents worth and maybe it will help??

Cheers,

Singtoh

[ewangr]

I have moved the network share to now mount under the /media folder. Figure if that is the problem, then at least I won't see the clocking in the home folder. Will see how it goes.