Hey! Fedora 22 is the first fedora release ever where all the services that use the network seem to start correctly! I haven't had to put any delayed restart commands in rc.local to get mail and NFS mounts and such working :-).
On reboot, it still almost always hangs for several minutes doing God knows what (since it kills the nvidia driver early, all I have is black screen, so I can't see what it might be doing).
I accidentally discovered that someone has at least made it listen to Ctrl-Alt-Delete during shutdown now. If I hold the keys down to make them repeat really fast systemd will stop waiting and go ahead and reboot.
Now if only I could figure out what it is waiting on (I suspect the "user daemons" for other users that are no longer logged in have something to do with it). Is there any way to have it trace shutdown and leave info I could examine the next time it is up?
What is with the tmpfs mounts it now makes for each user daemon? There is already a tmpfs /run, why does every user daemon need yet another mountpoint under that?
On 08/13/15 21:20, Tom Horsley wrote:
Hey! Fedora 22 is the first fedora release ever where all the services that use the network seem to start correctly! I haven't had to put any delayed restart commands in rc.local to get mail and NFS mounts and such working :-).
On reboot, it still almost always hangs for several minutes doing God knows what (since it kills the nvidia driver early, all I have is black screen, so I can't see what it might be doing).
I accidentally discovered that someone has at least made it listen to Ctrl-Alt-Delete during shutdown now. If I hold the keys down to make them repeat really fast systemd will stop waiting and go ahead and reboot.
Now if only I could figure out what it is waiting on (I suspect the "user daemons" for other users that are no longer logged in have something to do with it). Is there any way to have it trace shutdown and leave info I could examine the next time it is up?
If you remove "rhgb quiet" from the linux line when booting you should be able to monitor the shutdown processes.
I had a system "hang" on reboot/shutdown when using NFS over WiFi and did not know what was happening until I did that. That problem was BZ'd and fixed however.
On Thu, 13 Aug 2015 21:54:57 +0800 Ed Greshko wrote:
If you remove "rhgb quiet" from the linux line when booting you should be able to monitor the shutdown processes.
Normally yes, but I have a UHD monitor and an nvidia card that only works with the binary driver and something gets killed off really early in the shutdown that leaves me with a black screen (I already have rhgb and quiet turned off).
On 08/13/15 23:26, Tom Horsley wrote:
On Thu, 13 Aug 2015 21:54:57 +0800 Ed Greshko wrote:
If you remove "rhgb quiet" from the linux line when booting you should be able to monitor the shutdown processes.
Normally yes, but I have a UHD monitor and an nvidia card that only works with the binary driver and something gets killed off really early in the shutdown that leaves me with a black screen (I already have rhgb and quiet turned off).
I see.
If you know the time you rebooted and if you do "journaltcl -b -1 > somefile" and then examine somefile for oddities at that time, does anything jump out?
On Thu, 13 Aug 2015 23:37:36 +0800 Ed Greshko wrote:
If you know the time you rebooted and if you do "journaltcl -b -1 > somefile" and then examine somefile for oddities at that time, does anything jump out?
My best theory is that this bug still happens (it was only closed because it was against f20, not because anyone actually claimed to have fixed it).
https://bugzilla.redhat.com/show_bug.cgi?id=1088619
One of these days I'll get ambitious enough to enabled all the systemd debugging stuff documented here and see if I can see anything:
I also see a pause on shutdown or reboot that appears to be caused by a cifs mount. F22
On Thu, Aug 13, 2015 at 11:10 PM, Rahul Sundaram metherid@gmail.com wrote:
Hi
On Thu, Aug 13, 2015 at 9:20 AM, Tom Horsley wrote:
What is with the tmpfs mounts it now makes for each user daemon? There is already a tmpfs /run, why does every user daemon need yet another mountpoint under that?
IIRC because tmpfs doesn't have quota support
Rahul
-- users mailing list users@lists.fedoraproject.org To unsubscribe or change subscription options: https://admin.fedoraproject.org/mailman/listinfo/users Fedora Code of Conduct: http://fedoraproject.org/code-of-conduct Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines Have a question? Ask away: http://ask.fedoraproject.org
-
Ed Greshko -
Go Canes -
Rahul Sundaram -
Tom Horsley