Categories
Nginx

Nginx packages updated to 1.0.6

Packages of Nginx 1.0.6 for Debian 6.0 “Squeeze” and Debian 5.0 “Lenny” are now available in amd64 and i386 flavours.

This release brings faster cache loader, better https, lingering close compatibility and 9 bugfixes . Please take a look at the list of changes for more details.

Please note that :

21 replies on “Nginx packages updated to 1.0.6”

thanks for the package updates.

How about a user forum, so people can discuss stuff relating to dotdeb more freely, rather than having to post against specific blog entries? Would be useful I think, and allow people to help each other more.

It seems that the module nginx-http-push is part of the full package, right ?

I can see logs like: Using 32768KiB of shared memory for push module

How one is supposed to disable it ?

Thanks.

Hello Guillaume,

Thanks for nice packages. But I have little problem, if I restart nginx – it starts logging to ie. /var/log/nginx.log, but after log ratation /var/log/nginx.log becomes empty, and /var/log/nginx.log.1 is beaing created.. and further logging goes not to: /var/log/nginx.log
but to: /var/log/nginx.log.1…

I am not sure what can cause this problem.. nginx config, or some cron rotators do smth wrong..?

Thank you

Hello Guillaume, thanks for pointers,

Firstly I tried to edit users in this logrotate files, since my nginx runs under custom user.
So I waited other day – no luck.

Then I tried as root:
kill -USR1 `cat /var/run/nginx.pid`
kill -USR1 `pidof nginx`
but still no luck 🙁

Maybe you have any other ideas?

Thank you

Hey,

just tried this: kill -HUP `cat /var/run/nginx.pid` and it worked.

Isn’t it harmfull for nginx? I mean if the user will be visiting the site at the time of this command beeing run?

Thank you

I get this errors running in debian lenny:

2011/09/16 11:25:20 [emerg] 12106#0: bind() to 0.0.0.0:80 failed (98: Address already in use)
2011/09/16 11:25:20 [emerg] 12106#0: bind() to 0.0.0.0:443 failed (98: Address already in use)
2011/09/16 11:25:29 [alert] 12114#0: epoll_ctl(1, 0) failed (1: Operation not permitted)
2011/09/16 11:25:29 [alert] 12114#0: failed to register channel handler while initializing push module worker (1: Operation not permitted)
2011/09/16 11:25:29 [alert] 12112#0: cache manager process 12114 exited with fatal code 2 and can not be respawn
2011/09/16 11:25:29 [alert] 12115#0: epoll_ctl(1, 0) failed (1: Operation not permitted)
2011/09/16 11:25:29 [alert] 12115#0: failed to register channel handler while initializing push module worker (1: Operation not permitted)

Is this normal? My old 0.8.x build didn’t have such errors…

There is no other process running on port 80 or 443.

Furthermore the cache manager seems to be broken:

NEW system (1.0.6 from dotdeb):
ps aux | grep nginx
root 12244 0.0 0.0 73876 2488 ? Ss 12:04 0:00 nginx: master process /usr/sbin/nginx
www-data 12245 0.0 0.1 75964 6468 ? S 12:04 0:00 nginx: worker process

OLD system (self build 0.8.53):
www-data 12953 0.0 0.2 40780 8052 ? S Jul04 71:55 nginx: worker process
www-data 12954 0.0 0.1 38456 4308 ? S Jul04 0:15 nginx: cache manager process
root 29813 0.0 0.1 38456 4528 ? Ss Jun10 0:00 nginx: master process /usr/sbin/nginx

ok maybe you can build this packet without the http push module which seems to be broken at present…?

Again to log rotations, I like to put the logs of each host into its own sub-directory. Eg. /var/log/nginx/somedomain.com/access.log.

However it seems that files in subdirectories are ignored by /etc/logrotate.d/nginx.

Is this correct and how can I get around it?

Dubby

Comments are closed.