ping from iputils-20010805-alt2 (as well as from alt1) sometimes write this message on one of the freshest kernels (2.4.17). The following session log will show how it happens (nothing was chnaged in the system during this session): $ ping lrn.ru PING lrn.ru (62.118.250.6) from 172.30.180.16 : 56(84) bytes of data. 64 bytes from linux.ru.net (62.118.250.6): icmp_seq=1 ttl=247 time=13.196 msec 64 bytes from linux.ru.net (62.118.250.6): icmp_seq=2 ttl=247 time=13.813 msec 64 bytes from linux.ru.net (62.118.250.6): icmp_seq=3 ttl=247 time=17.530 msec 64 bytes from linux.ru.net (62.118.250.6): icmp_seq=4 ttl=247 time=11.357 msec --- lrn.ru ping statistics --- 4 packets transmitted, 4 received, 0% loss, time 3026ms rtt min/avg/max/mdev = 11.357/13.974/17.530/2.243 ms $ ping shamrock PING debialt.office.altlinux.ru (194.67.87.167) from 172.30.180.16 : 56(84) bytes of data. WARNING: kernel is not very fresh, upgrade is recommended. From 172.30.180.254: icmp_seq=198 Destination Host Unreachable From 172.30.180.254 icmp_seq=198 Destination Host Unreachable From MSK-M9-RBNet-1.RBNet.ru (195.209.14.153) icmp_seq=659 Time to live exceeded From MSK-M9-RBNet-1.RBNet.ru (195.209.14.153) icmp_seq=660 Time to live exceeded From MSK-M9-RBNet-1.RBNet.ru (195.209.14.153) icmp_seq=661 Time to live exceeded From ix-5-4.core1.Frankfurt2.Teleglobe.net (195.219.96.94) icmp_seq=662 Time to live exceeded From ix-5-4.core1.Frankfurt2.Teleglobe.net (195.219.96.94) icmp_seq=663 Time to live exceeded --- debialt.office.altlinux.ru ping statistics --- 663 packets transmitted, 0 received, +7 errors, 100% loss, time 662706ms , pipe 2 $ uname -a Linux arrakis.zephyrous 2.4.17-alt1-up #1 Wed Dec 26 19:07:36 MSK 2001 i686 unknown $ rpm -q iputils iputils-20010805-alt2 $ A long period of time (5-10 min) passed between the command \"ping shamrock\" was given and the message \"kernel is not very fresh\" appeared. In parallel, I started another \"ping shamrock\" and there was also a long pause, but no message about the kernel appeared! Here it is: $ ping shamrock PING debialt.office.altlinux.ru (194.67.87.167) from 172.30.180.16 : 56(84) bytes of data. From 172.30.180.254 icmp_seq=717 Destination Host Unreachable From MSK-M9-RBNet-1.RBNet.ru (195.209.14.153) icmp_seq=1178 Time to live exceeded From MSK-M9-RBNet-1.RBNet.ru (195.209.14.153) icmp_seq=1179 Time to live exceeded From MSK-M9-RBNet-1.RBNet.ru (195.209.14.153) icmp_seq=1180 Time to live exceeded From ix-5-4.core1.Frankfurt2.Teleglobe.net (195.219.96.94) icmp_seq=1181 Time to live exceeded From ix-5-4.core1.Frankfurt2.Teleglobe.net (195.219.96.94) icmp_seq=1182 Time to live exceeded From ix-5-4.core1.Frankfurt2.Teleglobe.net (195.219.96.94) icmp_seq=1183 Time to live exceeded --- debialt.office.altlinux.ru ping statistics --- 1184 packets transmitted, 0 received, +7 errors, 100% loss, time 1183643ms , pipe 2 $ --- ---
The attached tarball is a session log of straced ping where it reports not fresh kernel. It may help finding the problem.
А в чем собственно проблема?
Почему ping то говорит, что ядро не свежее, то не говорит этого. Может, это говорит о том, что есть какая-то ошибка внутри ping? Ядро довольно свежее: 2.4.17.
Понял, даже заглянул в iputils Там конечно без пол-литра не разберешься. Бага практически не отлавливаема Там есть две функции. Проверяется порядок их вызова. Если одна вдруг была вызвана раньше другой - то выдается предупреждение. Так что этот эффект: 1) ядерно-зависимый 2) плавающий 3) трудно вылавливаемый (мне не удалось на 2.4.18-ом получить тот же эффект)