postgresql - postmaster caused kernel hung issue -
postgresql - postmaster caused kernel hung issue -
i have 1 server has below kernel logs in /var/log/messages, looks postgres db process postmaster caused kernel hung issue, clues confirm that? , actions can prevent issue happen?
2014-10-22t05:35:29.140-05:00 localhost kernel: bug: bad page state in process postmaster pfn:42bfbd 2014-10-22t05:35:29.140-05:00 localhost kernel: page:ffffea000e99f158 flags:00c0000000000000 count:-1 mapcount:0 mapping:(null) index:20c53e (not tainted) 2014-10-22t05:35:29.140-05:00 localhost kernel: pid: 9543, comm: postmaster not tainted 2.6.32-358.23.2.el6.x86_64 #1 2014-10-22t05:35:29.140-05:00 localhost kernel: phone call trace: 2014-10-22t05:35:29.140-05:00 localhost kernel: [<ffffffff81128f37>] ? bad_page+0x107/0x160 2014-10-22t05:35:29.140-05:00 localhost kernel: [<ffffffff8112a73c>] ? get_page_from_freelist+0x72c/0x830 2014-10-22t05:35:29.140-05:00 localhost kernel: [<ffffffff8112bc43>] ? __alloc_pages_nodemask+0x113/0x8d0 2014-10-22t05:35:29.140-05:00 localhost kernel: [<ffffffff8112fe00>] ? __lru_cache_add+0x40/0x90 2014-10-22t05:35:29.140-05:00 localhost kernel: [<ffffffff81143c27>] ? handle_pte_fault+0x487/0xb50 2014-10-22t05:35:29.140-05:00 localhost kernel: [<ffffffff81160c6a>] ? alloc_pages_current+0xaa/0x110
this not postgres problem — it's kernel bug or hardware issue. utilize outdated (and insecure) kernel — yum update
server , reboot — if it's bug may corrected in latest kernel release , issue may go away.
postgresql linux-kernel kernel
Comments
Post a Comment