# By default does nginx just use a single log file? **URL:** \ **Category:** NGINX\ **Created:** [December 14, 2012, 9:01pm UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119 "2012-12-14T21:01:29Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![aris](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/a/258eb7/32.png) [@aris](https://www.ruby-forum.com/u/aris)\ **Post date:** [December 14, 2012, 9:01pm UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119/1 "2012-12-14T21:01:29Z")
By default does nginx just keep appending to the same log file? (for both access and error log files) Is there a reason why there is no built in way of having it create a new file per day/hour or something? I guess the idea is to keep the feature set very lean? ---
**Author:** ![S\_Ahmed](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/8c91f0/32.png) [@S\_Ahmed](https://www.ruby-forum.com/u/S_Ahmed)\ **Post date:** [December 14, 2012, 9:21pm UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119/2 "2012-12-14T21:21:21Z")
On Fri, Dec 14, 2012 at 03:00:53PM -0500, S Ahmed wrote: Hi there, > By default does nginx just keep appending to the same log file? (for both > access and error log files) Yes. (Although the access log file name can include per-request variables.) > Is there a reason why there is no built in way of having it create a new > file per day/hour or something? I guess the idea is to keep the feature > set very lean? Would you like it per day or per hour? Or every 4 MB, or every 1 million lines? (Or every second Tuesday unless the moon is full?) There are too many possibly-useful rules for anything built-in to satisfy everybody. So nginx allows you to choose whatever combination of circumstances you want, outside of nginx, and then provides a well-defined way for you to induce a log file rotation. > **[Controlling nginx](http://nginx.org/en/docs/control.html#logs)** > > nginx can be controlled with signals. The process ID of the master process is written to the file /usr/local/nginx/logs/nginx.pid by default. This name may be changed at configuration time, or in nginx.conf using the pid directive. The master process... That’s my understanding, anyway. ## f Francis D. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) ---
**Author:** ![S\_Ahmed](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/8c91f0/32.png) [@S\_Ahmed](https://www.ruby-forum.com/u/S_Ahmed)\ **Post date:** [December 14, 2012, 10:11pm UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119/3 "2012-12-14T22:11:01Z")
only on a full moon, excluding leap years. thank you and when can I expect this feature? 🙂 ---
**Author:** ![S\_Ahmed](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/8c91f0/32.png) [@S\_Ahmed](https://www.ruby-forum.com/u/S_Ahmed)\ **Post date:** [December 14, 2012, 10:43pm UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119/4 "2012-12-14T22:43:09Z")
Nginx doesn’t include periodic log rotation functionality, the only thing included is reopening the log files on USR1 signal, so one may rename the current log file and then kill -USR1 `cat /var/run/nginx.pid` for nginx to create a new log (see [Log Rotation | NGINX](http://wiki.nginx.org/LogRotation) ). External tools may be used to do the actual period rotation, like in Linuxish world it’s normally done by logrotate ( [http://manpages.ubuntu.com/manpages/lucid/man8/logrotate.8.html](http://manpages.ubuntu.com/manpages/lucid/man8/logrotate.8.html) ) - in recent Ubuntu versions it’s already set up to rotate default nginx log files. 2012/12/14 S Ahmed [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) ---
**Author:** ![S\_Ahmed](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/8c91f0/32.png) [@S\_Ahmed](https://www.ruby-forum.com/u/S_Ahmed)\ **Post date:** [December 15, 2012, 10:45am UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119/5 "2012-12-15T10:45:08Z")
On Dec 15, 2012, at 2:10 , S Ahmed wrote: > 1. kill -USR1 ‘cat /var/run/nginx.pid’ > 2. profit 1. rename file 2. kill -USR1 ‘cat /var/run/nginx.pid’ 3. sleep 2 4. do everything with the file: gzipping, moving to other host, etc. ---
**Author:** ![S\_Ahmed](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/8c91f0/32.png) [@S\_Ahmed](https://www.ruby-forum.com/u/S_Ahmed)\ **Post date:** [December 15, 2012, 7:20pm UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119/6 "2012-12-15T19:20:56Z")
I see, thanks. So there is there a small window of time when the logs will not be written or lost? ---
**Author:** ![S\_Ahmed](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/8c91f0/32.png) [@S\_Ahmed](https://www.ruby-forum.com/u/S_Ahmed)\ **Post date:** [December 14, 2012, 11:11pm UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119/7 "2012-12-14T23:11:12Z")
m irya, Thanks. I’m planning to run 12.04 ubuntu. What exactly does ‘re-opening’ of a log file mean? Isn’t it already open, meaning nginx has it open and is appending to it? So step by step, what is to be done? 1. kill -USR1 2. rename file? 3. kill -USR1 ‘cat /var/run/nginx.pid’ 4. profit ? ---
**Author:** ![S\_Ahmed](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/s/8c91f0/32.png) [@S\_Ahmed](https://www.ruby-forum.com/u/S_Ahmed)\ **Post date:** [December 15, 2012, 9:32pm UTC](https://www.ruby-forum.com/t/by-default-does-nginx-just-use-a-single-log-file/225119/8 "2012-12-15T21:32:20Z")
No, logs are always written either in renamed old file or created newly by master process. There is window between sending signal and reopening new files by worker processes. – Igor S. > **[Technical Support for NGINX and NGINX Plus Software](https://www.nginx.com/support/)** > > Get the help you need with support directly from the authors and maintainers of NGINX Open Source and NGINX Plus.