Re: Network gets tired, falls asleep

[Dan Williams <dcbw redhat com>]

Neal,

It appears here that the machine went to sleep?  The "errno = 19" reported  
by NetworkManager is ENODEV, and the telltale "remove event not handled", 
followed by the stuff about ACPI seem to indicate that the machine went to 
sleep for some reason, then woke up.

Right now, NM doesn't handle sleeping very well mainly due to the fact 
that we don't do anything iwth drivers when we sleep.  Most sleep scripts 
actually 'rmmod' the driver modules for their network cards, and re-insert 
them when the machine wakes up.  I think that is just a quick hack for 
drivers that don't properly support sleep.

If anyone has suggestions for making NM work better with ACPI/APM sleeps, 
I'd be quite open to them.

Dan

On Wed, 26 Jan 2005, Neal Becker wrote:

> Testing NM on FC3 using driverloader.  Seems to work for a while, but when I 
> left it on overnight it stopped working.  Here's some info:
> 
> Jan 25 20:40:47 nbecker2 kernel: eth1: New link status: Connected (0001)
> Jan 25 20:40:54 nbecker2 NetworkManager:   found! (encrypted) 
> Jan 25 20:40:54 nbecker2 NetworkManager:     SWITCH: best device changed
> Jan 25 20:40:54 nbecker2 NetworkManager: nm_state_modification_monitor(): 
> beginning activation for device 'eth1'
> Jan 25 20:40:55 nbecker2 NetworkManager: AP_NEED_KEY: access point is 
> encrypted, and a key exists.  No new key needed.
> Jan 25 20:40:59 nbecker2 NetworkManager: nm_device_set_wireless_config (eth1) 
> using essid 'nbecker', with Shared Key authentication.
> Jan 25 20:41:08 nbecker2 NetworkManager: Restarting mDNSResponder. 
> Jan 25 20:41:08 nbecker2 NetworkManager: nm_device_activation_worker(eth1): 
> device activated
> Jan 25 20:41:08 nbecker2 NetworkManager: nm_state_modification_monitor() 
> activated device eth1
> Jan 25 20:51:06 nbecker2 ntpd[2975]: sendto(195.197.174.39): Invalid argument
> Jan 25 20:51:08 nbecker2 ntpd[2975]: sendto(207.177.51.228): Invalid argument
> Jan 25 20:51:11 nbecker2 ntpd[2975]: sendto(206.222.212.234): Invalid argument
> Jan 25 20:59:33 nbecker2 kernel: usbcore: deregistering driver driverloader
> Jan 25 20:59:34 nbecker2 NetworkManager: nm_device_is_up() could not get flags 
> for device eth1.  errno = 19
> Jan 25 20:59:34 nbecker2 NetworkManager: nm_device_set_up_down() could not get 
> flags for device eth1.  errno = 19
> Jan 25 20:59:34 nbecker2 net.agent[11142]: remove event not handled
> Jan 25 20:59:34 nbecker2 net.agent[11150]: remove event not handled
> Jan 25 20:59:35 nbecker2 NetworkManager: nm_device_get_mode (eth1): error 
> setting card to Infrastructure mode.  errno = 19
> Jan 25 20:59:35 nbecker2 NetworkManager: nm_device_get_mode (eth1): error 
> setting card to Infrastructure mode.  errno = 19
> Jan 25 20:59:35 nbecker2 NetworkManager: nm_device_set_mode (eth1): error 
> setting card to Infrastructure mode.  errno = 19
> Jan 25 20:59:35 nbecker2 NetworkManager: nm_device_get_frequency(): error 
> getting frequency for device eth1.  errno = 19
> Jan 25 20:59:35 nbecker2 NetworkManager: nm_device_get_essid(): error getting 
> ESSID for device eth1.  errno = 19
> Jan 25 20:59:35 nbecker2 kernel: usbcore: registered new driver driverloader
> Jan 25 20:59:35 nbecker2 kernel: driverloader: bcmwl564.sys: 
> 0xffffff0001242000..0xffffff00012bb000 entry = ffffff0001249cd0 (pbase = 
> 0x10000)
> Jan 25 20:59:35 nbecker2 kernel: ACPI: PCI interrupt 0000:02:02.0[A] -> GSI 17 
> (level, low) -> IRQ 217
> Jan 25 20:59:35 nbecker2 kernel: 0000:02:02.0: cache line size not set; 
> forcing 32
> Jan 25 20:59:35 nbecker2 kernel: eth1: New link status: Disconnected (0002)
> Jan 25 20:59:37 nbecker2 kernel: ip_tables: (C) 2000-2002 Netfilter core team
> Jan 25 20:59:37 nbecker2 kernel: eth1: New link status: Connected (0001)
> _______________________________________________
> NetworkManager-list mailing list
> NetworkManager-list gnome org
> http://mail.gnome.org/mailman/listinfo/networkmanager-list
>