# Ngix Performace as a Reverse Proxy **URL:** \ **Category:** NGINX\ **Created:** [October 26, 2012, 6:37am UTC](https://www.ruby-forum.com/t/ngix-performace-as-a-reverse-proxy/223999 "2012-10-26T06:37:33Z")\ **Posts on this page:** 4\ **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:** [October 26, 2012, 6:37am UTC](https://www.ruby-forum.com/t/ngix-performace-as-a-reverse-proxy/223999/1 "2012-10-26T06:37:33Z")
Hi, We have been doing some testing with Nginx as a reverse proxy. We have been comparing it to a number of solutions which it easily beats, like IIS and Apache with mod\_proxy etc. However, as an experiment we have been comparing it to an adapted NIO server written in Java. This seems to be out performing Nginx in the reverse proxy role by a factor of 3 times. We are convinced our configuration is wrong. Both run on the same box (at different times) with the same sysctl settings (see below). We also saw some spikes, up to 3 seconds per request at times, and some at 10 over a 1 million request test of 1000 concurrent clients. We are using a fairly straight forward configuration for Nginx. Since we have two processors on the box we tried worker\_processes of 4 with worker\_connections of 6000, then we tried worker\_processes of 40 with worker\_connections of 5000. No change. We need to be able to support responsive Ajax requests with strategies like HTTP streaming and long polling in our setup. Any ideas what we can do to boost our throughput and latency? [root@dc1dmzngx02 apachebench]# uname -a Linux dc1dmzngx02 2.6.32-220.13.1.el6.x86\_64 #1 SMP Tue Apr 17 23:56:34 BST 2012 x86\_64 x86\_64 x86\_64 GNU/Linux [root@dc1dmzngx02 apachebench]# cat /etc/sysctl.conf # Kernel sysctl configuration file for Red Hat Linux # # For binary values, 0 is disabled, 1 is enabled. See sysctl(8) and # sysctl.conf(5) for more details. # High perf config net.core.somaxconn = 12048 net.core.rmem\_default = 262144 net.core.wmem\_default = 262144 net.core.rmem\_max = 16777216 net.core.wmem\_max = 16777216 net.ipv4.tcp\_rmem = 4096 4096 16777216 net.ipv4.tcp\_wmem = 4096 4096 16777216 net.ipv4.tcp\_mem = 786432 2097152 3145728 net.ipv4.tcp\_max\_syn\_backlog = 16384 net.core.netdev\_max\_backlog = 20000 net.ipv4.tcp\_fin\_timeout = 15 net.ipv4.tcp\_max\_syn\_backlog = 16384 net.ipv4.tcp\_tw\_reuse = 1 net.ipv4.tcp\_tw\_recycle = 1 net.ipv4.tcp\_max\_orphans = 131072 # Controls IP packet forwarding net.ipv4.ip\_forward = 0 # Controls source route verification net.ipv4.conf.default.rp\_filter = 1 # Do not accept source routing net.ipv4.conf.default.accept\_source\_route = 0 # Controls the System Request debugging functionality of the kernel kernel.sysrq = 0 # Controls whether core dumps will append the PID to the core filename. # Useful for debugging multi-threaded applications. kernel.core\_uses\_pid = 1 # Controls the use of TCP syncookies net.ipv4.tcp\_syncookies = 1 # Disable netfilter on bridges. net.bridge.bridge-nf-call-ip6tables = 0 net.bridge.bridge-nf-call-iptables = 0 net.bridge.bridge-nf-call-arptables = 0 # Controls the maximum size of a message, in bytes kernel.msgmnb = 65536 # Controls the default maxmimum size of a mesage queue kernel.msgmax = 65536 # Controls the maximum shared segment size, in bytes kernel.shmmax = 68719476736 # Controls the maximum number of shared memory segments, in pages kernel.shmall = 4294967296 Thanks, Niall ---
**Author:** ![Niall\_GallagherSY](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/n/a5b964/32.png) [@Niall\_GallagherSY](https://www.ruby-forum.com/u/Niall_GallagherSY)\ **Post date:** [October 26, 2012, 2:05pm UTC](https://www.ruby-forum.com/t/ngix-performace-as-a-reverse-proxy/223999/2 "2012-10-26T14:05:03Z")
On 26 Oct2012, at 08:36 , Niall Gallagher - Yieldbroker [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > Hi, > > We have been doing some testing with Nginx as a reverse proxy. We have been > comparing it to a number of solutions which it easily beats, like IIS and Apache > with mod\_proxy etc. However, as an experiment we have been comparing it to an > adapted NIO server written in Java. This seems to be out performing Nginx in the > reverse proxy role by a factor of 3 times. We are convinced our configuration is > wrong. Both run on the same box (at different times) with the same sysctl settings > (see below). We also saw some spikes, up to 3 seconds per request at times, and > some at 10 over a 1 million request test of 1000 concurrent clients. > > We are using a fairly straight forward configuration for Nginx. Since we have > two processors on the box we tried worker\_processes of 4 with worker\_connections > of 6000, then we tried worker\_processes of 40 with worker\_connections of 5000. No > change. We need to be able to support responsive Ajax requests with strategies > like HTTP streaming and long polling in our setup. > > Any ideas what we can do to boost our throughput and latency? Buffers. nginx should not write/read anything to/from disk while proxing for maximum performance. Check your average request and response sizes and tune buffers sizes accordingly (look at proxy\_buffers, proxy\_buffer\_size, proxy\_max\_temp\_file\_size documentation). ---
**Author:** ![Niall\_GallagherSY](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/n/a5b964/32.png) [@Niall\_GallagherSY](https://www.ruby-forum.com/u/Niall_GallagherSY)\ **Post date:** [October 28, 2012, 11:26pm UTC](https://www.ruby-forum.com/t/ngix-performace-as-a-reverse-proxy/223999/3 "2012-10-28T23:26:46Z")
According to documentation for proxy\_buffering “For Comet applications based on long-polling it is important to set proxy\_buffering to off, otherwise the asynchronous response is buffered and the Comet does not work.” I have tried the following and am not getting better results, however its still being outperformed by the Java HTTP proxy by about 10% - 15%. worker\_processes 2; worker\_cpu\_affinity 01 10; worker\_rlimit\_nofile 20000; However spikes of up to 20 seconds are still frequent under high load. ---
**Author:** ![Niall\_GallagherSY](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/n/a5b964/32.png) [@Niall\_GallagherSY](https://www.ruby-forum.com/u/Niall_GallagherSY)\ **Post date:** [October 29, 2012, 10:12am UTC](https://www.ruby-forum.com/t/ngix-performace-as-a-reverse-proxy/223999/4 "2012-10-29T10:12:54Z")
On 29 Oct2012, at 02:26 , Niall Gallagher - Yieldbroker [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid) wrote: > According to documentation for proxy\_buffering > > “For Comet applications based on long-polling it is important to set > proxy\_buffering to off, otherwise the asynchronous response is buffered and the > Comet does not work.” > I have tried the following and am not getting better results, however its still > being outperformed by the Java HTTP proxy by about 10% - 15%. Try to set postpone\_output 0;