Messages in this thread |  | | From | (Danny ter Haar) | Subject | 2.0.27: kernel weirdness | Date | 25 Jan 1997 09:10:27 +0100 |
| |
P5-100Mhz 32Mb 1.2Gb EIDE disk WD8013 Ethernet card Asus mainboard Debian 1.1 This machine is basically our nameserver, NIS server and runs msql and apache (for management functions only)
Linux 2.0.27 (dth@troi) (gcc 2.7.2) #1 Thu Dec 5 22:50:28 MET 1996 [troi]
Memory: Total Used Free Shared Buffers Cached Mem: 31024 29744 1280 8868 13900 8140 Swap: 40316 1564 38752
Bootup: Sun Dec 15 10:55:12 1996 Load average: 0.08 0.05 0.01 2/40 4558
user : 20:32:06.70 2.1% page in : 15460478 disk 1: 6560500r 5406173w nice : 0:14:49.46 0.0% page out: 7136074 system: 19:44:41.09 2.0% swap in : 2713624 idle : 39d 5:40:08.14 95.9% swap out: 11260 uptime: 40d 22:11:45.37 context : 138623495
irq 0: 353590539 timer irq 8: 0 irq 1: 780 keyboard irq 9: 0 irq 2: 0 cascade irq 10: 627502185 WD8013 irq 3: 12954 + serial irq 11: 0 irq 4: 6 irq 12: 0 irq 5: 0 irq 13: 1 math error irq 6: 2 irq 14: 61542371 + ide0 irq 7: 0 irq 15: 0
Twice within the current uptime we had this kernel message:
free_one_pmd: bad directory entry 80000000
At this point certain processes get killed. Yesterday it was bind, previous it was apache and NIS. Just restarting the "missing" processes makes everything work like normal. Prior uptime (2.0.23 kernel) was more than 60 days. Sudden death (probably ping_of_death) made us install 2.0.27.
Anyone some ideas ?
Danny -- Danny ter Haar | Cistron Internet Services | Unix & Internet dth@het.net | finger dth@cistron.nl for PGP-key | specialists ---------------------------+31-172-419445--------------------SP6------
|  |