Opened 12 years ago
Closed 12 years ago
#356 closed defect (worksforme)
$request_time and $upstream_response_time exactly the same since 1.3.15
| Reported by: | Mathieu Martin | Owned by: | |
|---|---|---|---|
| Priority: | minor | Milestone: | |
| Component: | nginx-core | Version: | 1.3.x |
| Keywords: | upstream, time | Cc: | |
| uname -a: | Linux (removed) 2.6.32-358.6.1.el6.x86_64 #1 SMP Tue Apr 23 19:29:00 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux | ||
| nginx -V: |
nginx version: nginx/1.4.1
built by gcc 4.4.7 20120313 (Red Hat 4.4.7-3) (GCC) TLS SNI support enabled configure arguments: --prefix=/opt/nginx-1.4.1 --conf-path=/etc/nginx/nginx.conf --sbin-path=/opt/nginx-1.4.1/sbin/nginx --with-http_ssl_module --with-http_gzip_static_module --with-http_realip_module |
||
Description
Since deploying nginx 1.4.1, we've noticed that $request_time and $upstream_response_time are exactly the same (for requests served with the help of an upstream server). We've observed this situation on hundreds of requests.
I'm seeing this on CentOS 6.4, installing from source.
I've checked a few versions.
The following versions are fine (the times are slightly different): 1.2.7, 1.2.8, 1.2.9, 1.3.10, 1.3.12, 1.3.14
The following versions have the problem (the times are precisely the same): 1.3.15, 1.4.0, 1.4.1
This is my first time reporting a bug here. Please let me know if you want me to clean up my bug report, or if anything's amiss.
Thanks for the awesome work.
Mat
Note:
See TracTickets
for help on using tickets.
As of 1.3.15, $request_time calculation was slightly changed due to some optimizations introduced, and for a first request in a connection it no longer includes time between opening a connection and reading a first byte from a client. It still works correctly though. With the following configuration:
log_format foo "request_time:$request_time " "upstream_response_time:$upstream_response_time"; access_log /path/to/log foo;I see the following in logs by typing a request by hand with telnet: