errors

[Alexey Bestciokov <proforg@xxxxxxxxxxxx>]

Игорь, success srory в очередной раз не получилось :)
после очередного перезапуска nginx стал вести себя весьма странно,
судя по логу воркеры постоянно умирают
2005/01/30 00:56:34 [notice] 20491#0: signal 17 (SIGCHLD) received
2005/01/30 00:56:34 [alert] 20491#0: worker process 21057 exited on signal 11
2005/01/30 00:56:34 [notice] 20491#0: start worker process 21071
2005/01/30 00:56:34 [notice] 20491#0: signal 29 (SIGIO) received
2005/01/30 00:56:37 [notice] 20491#0: signal 17 (SIGCHLD) received
2005/01/30 00:56:37 [alert] 20491#0: worker process 21065 exited on signal 11
2005/01/30 00:56:37 [notice] 20491#0: start worker process 21074
2005/01/30 00:56:37 [notice] 20491#0: signal 29 (SIGIO) received
2005/01/30 00:56:40 [notice] 20491#0: signal 17 (SIGCHLD) received
2005/01/30 00:56:40 [alert] 20491#0: worker process 21071 exited on signal 11
2005/01/30 00:56:40 [notice] 20491#0: start worker process 21075
2005/01/30 00:56:40 [notice] 20491#0: signal 29 (SIGIO) received
2005/01/30 00:56:43 [notice] 20491#0: signal 17 (SIGCHLD) received
2005/01/30 00:56:43 [alert] 20491#0: worker process 21074 exited on signal 11
2005/01/30 00:56:43 [notice] 20491#0: start worker process 21076
2005/01/30 00:56:43 [notice] 20491#0: signal 29 (SIGIO) received

при этом апач за ним ведёт себя так словно понятия keepalive для него не
существует ...
при попытке что либо скачать - скачиваются первые 100Кб-2мб после чего
соединение закрывается.
Может я в конфиге чего напутал ? хотя вроде всё проверил ...

nginx 0.16
linux 2.6.10

конфиг:

worker_processes  10;
debug_points  abort;
events {
    connections  2048;
    use epoll;
    epoll_events 1024;
}

http {
    sendfile  on;
    tcp_nopush on;
    tcp_nodelay on;
    ....
    }

Алексей Бещёков.
proforg@xxxxxxxxxxxx