Our new server (Samba File Server) running redhat enterprise 2.1 croaked over the weekend. I noted lots of kernel errors like the following:
Jun 8 07:13:42 sv1 kernel: Unable to handle kernel paging request at virtual address ed491850 Jun 8 07:13:42 sv1 kernel: printing eip: Jun 8 07:13:42 sv1 kernel: c013025f Jun 8 07:13:42 sv1 kernel: *pde = 00000000 Jun 8 07:13:42 sv1 kernel: Oops: 0000 Jun 8 07:13:42 sv1 kernel: sd_mod r128 agpgart autofs 3c59x usb-storage scsi_mod usb-uhci ehci-hcd usbcore ext3 jbd raid1 Jun 8 07:13:42 sv1 kernel: CPU: 0 Jun 8 07:13:42 sv1 kernel: EIP: 0010:[kmem_cache_alloc+127/224] Not tainted Jun 8 07:13:42 sv1 kernel: EIP: 0010:[<c013025f>] Not tainted Jun 8 07:13:42 sv1 kernel: EFLAGS: 00010082 Jun 8 07:13:42 sv1 kernel: Jun 8 07:13:42 sv1 kernel: EIP is at kmem_cache_alloc [kernel] 0x7f (2.4.20-13.7) Jun 8 07:13:42 sv1 kernel: eax: 0a06060e ebx: c25a616c ecx: c5310000 edx: 00000080 Jun 8 07:13:42 sv1 kernel: esi: 00000246 edi: 0a06060e ebp: dbe8ba05 esp: d3a65ce8 Jun 8 07:13:42 sv1 kernel: ds: 0018 es: 0018 ss: 0018 Jun 8 07:13:42 sv1 kernel: Process rhn_check (pid: 29442, stackpage=d3a65000) Jun 8 07:13:42 sv1 kernel: Stack: e0813211 dcb19480 dcb19480 d3a65d08 c25a616c fffffff4 de0dfa80 deba3300 Jun 8 07:13:42 sv1 kernel: c014f0dc c25a616c 000001f0 dff31bd0 dbe7600a 3b6ee9f0 0000000d d3a65eb0 Jun 8 07:13:42 sv1 kernel: fffffff4 de0dfa80 deba3300 c01469dc deba3300 d3a65eb0 deba3300 00000000 Jun 8 07:13:42 sv1 kernel: Call Trace: [3c59x:__insmod_3c59x_O/lib/modules/2.4.20-13.7/kernel/drivers/net+-990703/96] jou rnal_dirty_metadata_R0931dd6e [jbd] 0x151 (0xd3a65ce8)) Jun 8 07:13:42 sv1 kernel: Call Trace: [<e0813211>] journal_dirty_metadata_R0931dd6e [jbd] 0x151 (0xd3a65ce8)) Jun 8 07:13:42 sv1 kernel: [d_alloc+28/384] d_alloc [kernel] 0x1c (0xd3a65d08)) Jun 8 07:13:42 sv1 kernel: [<c014f0dc>] d_alloc [kernel] 0x1c (0xd3a65d08)) Jun 8 07:13:42 sv1 kernel: [real_lookup+60/192] real_lookup [kernel] 0x3c (0xd3a65d34)) Jun 8 07:13:42 sv1 kernel: [<c01469dc>] real_lookup [kernel] 0x3c (0xd3a65d34)) Jun 8 07:13:42 sv1 kernel: [link_path_walk+1658/2304] link_path_walk [kernel] 0x67a (0xd3a65d50)) Jun 8 07:13:42 sv1 kernel: [<c01471da>] link_path_walk [kernel] 0x67a (0xd3a65d50)) Jun 8 07:13:42 sv1 kernel: [path_release+16/48] path_release [kernel] 0x10 (0xd3a65d8c)) Jun 8 07:13:42 sv1 kernel: [<c0146930>] path_release [kernel] 0x10 (0xd3a65d8c)) Jun 8 07:13:42 sv1 kernel: [link_path_walk+2286/2304] link_path_walk [kernel] 0x8ee (0xd3a65da8)) Jun 8 07:13:42 sv1 kernel: [<c014744e>] link_path_walk [kernel] 0x8ee (0xd3a65da8)) Jun 8 07:13:42 sv1 kernel: [3c59x:__insmod_3c59x_O/lib/modules/2.4.20-13.7/kernel/drivers/net+-992072/96] journal_get_write _access_Rc2c3fc04 [jbd] 0x38 (0xd3a65dd0)) Jun 8 07:13:42 sv1 kernel: [<e0812cb8>] journal_get_write_access_Rc2c3fc04 [jbd] 0x38 (0xd3a65dd0)) Jun 8 07:13:42 sv1 kernel: [__alloc_pages+123/848] __alloc_pages [kernel] 0x7b (0xd3a65e10)) Jun 8 07:13:42 sv1 kernel: [<c013627b>] __alloc_pages [kernel] 0x7b (0xd3a65e10)) Jun 8 07:13:42 sv1 kernel: [__pte_chain_free+22/32] __pte_chain_free [kernel] 0x16 (0xd3a65e2c)) Jun 8 07:13:42 sv1 kernel: [<c013aa46>] __pte_chain_free [kernel] 0x16 (0xd3a65e2c)) Jun 8 07:13:42 sv1 kernel: [do_anonymous_page+519/544] do_anonymous_page [kernel] 0x207 (0xd3a65e38)) Jun 8 07:13:42 sv1 kernel: [<c01281a7>] do_anonymous_page [kernel] 0x207 (0xd3a65e38)) Jun 8 07:13:42 sv1 kernel: [3c59x:__insmod_3c59x_O/lib/modules/2.4.20-13.7/kernel/drivers/net+-917243/96] ext3_mark_iloc_di rty [ext3] 0x35 (0xd3a65e40)) Jun 8 07:13:42 sv1 kernel: [<e0825105>] ext3_mark_iloc_dirty [ext3] 0x35 (0xd3a65e40)) Jun 8 07:13:42 sv1 kernel: [do_no_page+60/624] do_no_page [kernel] 0x3c (0xd3a65e60)) Jun 8 07:13:42 sv1 kernel: [<c01281fc>] do_no_page [kernel] 0x3c (0xd3a65e60)) It almost looks like the network card (a 3c905) might be part of this. I've used lots of 3c905 with no problems, but it is new hardware & could concievably have some problems. The servers been running for a week with no problems up until this weekend. The server has few departures from the stock enterprise setup including qmail installed, 24.20 kernel (needed usb2 support), and webmin. Any ideas anyone? Thanks, -- Neil Jolly (with Yoda-like voice) "Confrontation leads to anger... Anger leads to fear... Fear leads to using Windows NT in mission-critical combat systems... And this is how the ancients fell...
