Hi Dennis, Keith,
Thanks for your pointers. I tried disabling C-states and and pcie-aspm options. No luck so far. My box has to be rebooted once in every 2-3 days. Following is the last info in my syslog before reboot.
Jan 21 01:33:41 usmovd dnsmasq-dhcp[3254]: DHCPREQUEST(virbr0) 192.168.122.171 00:e2:7e:31:d2:8b Jan 21 01:33:41 usmovd dnsmasq-dhcp[3254]: DHCPACK(virbr0) 192.168.122.171 00:e2:7e:31:d2:8b Jan 21 01:52:44 usmovd dnsmasq-dhcp[3254]: DHCPREQUEST(virbr0) 192.168.122.194 00:c6:ab:4c:20:71 Jan 21 01:52:44 usmovd dnsmasq-dhcp[3254]: DHCPACK(virbr0) 192.168.122.194 00:c6:ab:4c:20:71 Jan 21 01:53:18 usmovd dhclient[2526]: DHCPREQUEST on eth0 to 135.227.220.25 port 67 (xid=0x3c4c435c) Jan 21 01:53:18 usmovd dhclient[2526]: DHCPACK from 135.227.220.25 (xid=0x3c4c435c) Jan 21 01:53:18 usmovd dhclient[2526]: bound to 135.227.220.131 -- renewal in 1595 seconds. Jan 21 01:53:18 usmovd NetworkManager[2495]: <info> (eth0): DHCPv4 state changed renew -> renew Jan 21 01:53:18 usmovd NetworkManager[2495]: <info> address 135.227.220.131 Jan 21 01:53:18 usmovd NetworkManager[2495]: <info> prefix 24 (255.255.255.0) Jan 21 01:53:18 usmovd NetworkManager[2495]: <info> gateway 135.227.220.1 Jan 21 01:53:18 usmovd NetworkManager[2495]: <info> nameserver '128.251.10.25' Jan 21 01:53:18 usmovd NetworkManager[2495]: <info> domain name ' eng.timetra.com' Jan 21 01:58:06 usmovd dnsmasq-dhcp[3254]: DHCPREQUEST(virbr0) 192.168.122.171 00:e2:7e:31:d2:8b Jan 21 01:58:06 usmovd dnsmasq-dhcp[3254]: DHCPACK(virbr0) 192.168.122.171 00:e2:7e:31:d2:8b <reboot> Jan 21 10:27:55 usmovd kernel: imklog 5.8.10, log source = /proc/kmsg started. Jan 21 10:27:55 usmovd rsyslogd: [origin software="rsyslogd" swVersion="5.8.10" x-pid="2317" x-info="http://www.rsyslog.com"] start <snip> Jan 21 10:27:55 usmovd kernel: Kernel command line: ro root=/dev/mapper/vg_usmovd-lv_root rd_NO_LUKS LANG=en_US.UTF-8 rd_NO_MD rd_LVM_LV=vg_usmovd/lv_swap SYSFONT=latarcyrheb-sun16 crashkernel=128M rd_LVM_LV=vg_usmovd/lv_root KEYBOARDTYPE=pc KEYTABLE=us rd_NO_DM rhgb quiet pcie_aspm.policy=off intel_idle.max_cstate=2
Thanks, --Arvind
Today's Topics:
4. Hangs after inactivity (Arvind Nagarajan) 5. Re: Hangs after inactivity (Keith Keller) 6. Re: Hangs after inactivity (Dennis Jacobfeuerborn)
Message: 4 Date: Mon, 30 Dec 2013 18:00:13 -0800 From: Arvind Nagarajan arnagara@gmail.com Subject: [CentOS] Hangs after inactivity To: centos@centos.org Message-ID: CAC9tjoHy2=_9KSLxj2pLcepWn9NmA3Sq3evr9-CK4DSF5G1vtw@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1
Hi,
My linux machine hangs after some period of inactivity (8+ hrs). Anyone experiencing similar issue.
CentOS Release 6.5(Final) Kernel Linux 2.6.32-431.el6_x86_64 GNOME 2.28.2
Processor (0 - 7): Intel(R) Core(TM) i7-4770 CPU @ 3.40 GHz
Any pointers will be helpful.
Thanks, --Arvind
------------------------------
Message: 5 Date: Mon, 30 Dec 2013 18:29:11 -0800 From: Keith Keller kkeller@wombat.san-francisco.ca.us Subject: Re: [CentOS] Hangs after inactivity To: centos@centos.org Message-ID: nom9paxiik.ln2@goaway.wombat.san-francisco.ca.us Content-Type: text/plain; charset=us-ascii
On 2013-12-31, Arvind Nagarajan arnagara@gmail.com wrote:
My linux machine hangs after some period of inactivity (8+ hrs). Anyone experiencing similar issue.
[snip]
Any pointers will be helpful.
The canonical first place to check is your logs, many of which are in /var/log/ . Beyond that, information about running apps at the time you leave it before it crashes would be helpful, and you might try not logging in to see if the problem occurs with no user activity (if not, it points to an app that starts when you log in).
--keith
-- kkeller@wombat.san-francisco.ca.us
------------------------------
Message: 6 Date: Tue, 31 Dec 2013 05:08:35 +0100 From: Dennis Jacobfeuerborn dennisml@conversis.de Subject: Re: [CentOS] Hangs after inactivity To: centos@centos.org Message-ID: 52C24343.2070208@conversis.de Content-Type: text/plain; charset=ISO-8859-1; format=flowed
On 31.12.2013 03:00, Arvind Nagarajan wrote:
Hi,
My linux machine hangs after some period of inactivity (8+ hrs). Anyone experiencing similar issue.
CentOS Release 6.5(Final) Kernel Linux 2.6.32-431.el6_x86_64 GNOME 2.28.2
Processor (0 - 7): Intel(R) Core(TM) i7-4770 CPU @ 3.40 GHz
Any pointers will be helpful.
You can try running the machine with the kernel option "pcie_aspm=off" and see if that helps and also try to disable C-States or Power Saving all together in the BIOS. More Details here:
https://access.redhat.com/site/documentation/en-US/Red_Hat_Enterprise_Linux/...
http://lists.us.dell.com/pipermail/linux-poweredge/2011-June/044901.html
I've experienced lock-ups in the past because of these issues. The network device hung because it tried to go into power saving mode using ASPM and it failed and whole servers locked-up after being idle for a while because of a bug in Intel Processors regarding the deeper power saving modes (C-States) although these are supposed to be fixed since 6.3. Hope this helps.
Regards, Dennis