Search Results

Search found 47 results on 2 pages for 'gunicorn'.

Page 2/2 | < Previous Page | 1 2 

• nginx reverse proxy slows down my throughput by half

  - by Isaac A Mosquera

  I'm currently using nginx to proxy back to gunicorn with 8 workers. I'm using an amazon extra large instance with 4 virtual cores. When I connect to gunicorn directly I get about 10K requests/sec. When I serve a static file from nginx I get about 25 requests/sec. But when I place gunicorn behind nginx on the same physical server I get about 5K requests/sec. I understand there will be some latency from nginx, but I think there might be a problem since it's a 50% drops. Anybody heard of something similar? any help would be great! Here is the relevant nginx conf: worker_processes 4; worker_rlimit_nofile 30000; events { worker_connections 5120; } http { sendfile on; tcp_nopush on; tcp_nodelay on; keepalive_timeout 65; types_hash_max_size 2048; } sites-enabled/default: upstream backend { server 127.0.0.1:8000; } server { server_name api.domain.com ; location / { proxy_pass http://backend; proxy_buffering off; } }

  Read the article

• Limit which processes a user can restart with supervisor?

  - by dvcolgan

  I have used supervisor to manage a Gunicorn process running a Django site, though this question could pertain to anything being managed by supervisor. Previously I was the only person managing and using our server, and supervisor just ran as root and I would use sudo to run supervisorctl restart myapp when needed. Now our server has to support multiple users working on different sites, and each project needs to be able to restart their own gunicorn processes without being able to restart other users' processes. I followed this blog post: http://drumcoder.co.uk/blog/2010/nov/24/running-supervisorctl-non-root/ and was able to allow non-root users to use supervisorctl, but now anyone can restart anyone else's processes. From the looks of it, supervisor doesn't have a way of doing per-user access control. Anyone have any ideas on how to allow users to restart only their own processes without root?

  Read the article

• Varnish server in front of nginx server with multiple virtualhosts

  - by Garreth 00

  I have tried to search for a solution for this, but can't find any documentation/tips on my specific setup. My setup: Backendserver: ngnix: 2 different websites (2 top domains) in virtualenv, running gunicorn/python/django Backendserver hardware(VPS) 2gb ram, 8 CPU Databaseserver: postgresql - pg_bouncer Backendserver hardware (VPS) 1gb ram, 8 CPU Varnishserver: only running varnish Varnishserver hardware (VPS) 1gb ram, 8 CPU I'm trying to set up a varnish server to handle rare spike in traffic (20 000 unique req/s) The spike happens when a tv program mention one of the sites. What do I need to do, to make the varnish server cache both sites/domains on my backendserver? My /etc/varnish/default.vcl : backend django_backend { .host = "local.backendserver.com"; .port = "8080"; } My /usr/local/nginx/site-avaible/domain1.com upstream gunicorn_domain1 { server unix:/home/<USER>/.virtualenvs/<DOMAIN1>/<APP1>/run/gunicorn.sock fail_timeout=0; } server { listen 80; listen 8080; server_name domain1.com; rewrite ^ http://www.domains.com$request_uri? permanent; } server { listen 80 default_server; listen 8080; client_max_body_size 4G; server_name www.domain1.com; keepalive_timeout 5; # path for static files root /home/<USER>/<APP>-media/; location / { proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; proxy_set_header Host $http_host; proxy_redirect off; if (!-f $request_filename) { proxy_pass http://gunicorn_domain1; break; } } } My /usr/local/nginx/site-avaible/domain2.com upstream gunicorn_domain2 { server unix:/home/<USER>/.virtualenvs/<DOMAIN2>/<APP2>/run/gunicorn.sock fail_timeout=0; } server { listen 80; listen 8080; server_name domain2.com; rewrite ^ http://www.domains.com$request_uri? permanent; } server { listen 80; listen 8080; client_max_body_size 4G; server_name www.domain2.com; keepalive_timeout 5; # path for static files root /home/<USER>/<APP>-media/; location / { proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; proxy_set_header Host $http_host; proxy_redirect off; if (!-f $request_filename) { proxy_pass http://gunicorn_domain2; break; } } } Right now, If I try the Ip of the varnishserver I only get served domain1.com. Would everything be correct if I change the DNS of the two domain to point to the varnishserver, or is there extra setup before it would work? Question 2: Do I need a dedicated server for varnish, or could I just install varnish on my backendserver, or would the server run out of memory quick?

  Read the article

• Started an application through SSH, command line now gone, what happens next?

  - by Chris Dutrow

  Context: This is a very basic question Using Putty and SSH for the first time to do some serious server setup and run into the situation where I have started a process that I do not want to stop. The process is the gunicorn WSGI HTTP Server (running on Centos 6.3). The command I used to start the process is (as per their Quick Start): gunicorn -w 4 myapp:app At this point in the work session, I have lost the command prompt. This must be such a non-issue that it doesn't even enter into an experienced user's consciousness. But unfortunately at my level of experience, I am left with several fundamental questions: Does the fact that I have lost the command prompt mean that the process is still running? How do I get back to the command prompt without killing the process? How do I come back and monitor the process later? How do I eventually kill the process? Any help is appreciated, thanks so much!

  Read the article

• Prevent nginx from redirecting traffic from https to http when used as a reverse proxy

  - by Chris Pratt

  Here's my abbreviated nginx vhost conf: upstream gunicorn { server 127.0.0.1:8080 fail_timeout=0; } server { listen 80; listen 443 ssl; server_name domain.com ~^.+\.domain\.com$; location / { try_files $uri @proxy; } location @proxy { proxy_pass_header Server; proxy_redirect off; proxy_set_header Host $http_host; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; proxy_set_header X-Forwarded-Proto https; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Scheme $scheme; proxy_connect_timeout 10; proxy_read_timeout 120; proxy_pass http://gunicorn; } } The same server needs to serve both HTTP and HTTPS, however, when the upstream issues a redirect (for instance, after a form is processed), all HTTPS requests are redirected to HTTP. The only thing I have found that will correct this issue is changing proxy_redirect to the following: proxy_redirect http:// https://; That works wonderfully for requests coming from HTTPS, but if a redirect is issued over HTTP it also redirects that to HTTPS, which is a problem. Out of desperation, I tried: if ($scheme = 'https') { proxy_redirect http:// https://; } But nginx complains that proxy_redirect isn't allowed here. The only other option I can think of is to define the two servers separately and set proxy_redirect only on the SSL one, but then I would have duplicate the rest of the conf (there's a lot in the server directive that I omitted for simplicity sake). I know I could also use an include directive to factor out the redundancy, but I really want to keep just one conf file without any dependencies. So, first, is there something I'm missing that will negate the problem entirely? Or, second, if not, is there any other way (besides including an external file) to factor out the redundant config information so that I can separate out the HTTP and HTTPS versions of the server config?

  Read the article

• Arch Linux with an nginx/django setup refuses to display ANYTHING

  - by Holland

  I'm on Amazon Ec2, with an Arch Linux server. While I truly am loving it, I'm having the issue of actually getting nginx to display anything. Everytime I try to throw my hostname into the browser, the browser states that it's not available for some reason - almost as if the host doesn't even exist. One thing I'd like to know is, how can I get this up and running? Is there a specific arch linux configuration I have to do to make it web accessible? I have port 80 open, as well as port 22. I've tried using gunicorn, python-flup, and nginx. Nginx Config user http; worker_processes 1; #error_log logs/error.log; #error_log logs/error.log notice; #error_log logs/error.log info; pid logs/nginx.pid; events { worker_connections 1024; } http { include mime.types; default_type application/octet-stream; log_format main '$remote_addr - $remote_user [$time_local] "$request" ' '$status $body_bytes_sent "$http_referer" ' '"$http_user_agent" "$http_x_forwarded_for"'; access_log logs/access.log main; sendfile on; #tcp_nopush on; #keepalive_timeout 0; keepalive_timeout 65; #gzip on; server { listen 80; server_name _; access_log /var/log/nginx/access.log; error_log /var/log/nginx/error.log; #charset koi8-r; location ^~ /media/ { root /path/to/media; } location ^~ /admin-media/ { root /usr/lib/python2.7/site-packages/django/contrib/admin/media; } location / { root /path/to/root/; fastcgi_pass 127.0.0.1:8080; fastcgi_param SERVER_NAME $server_name; fastcgi_param SERVER_PORT $server_port; fastcgi_param SERVER_PROTOCOL $server_protocol; fastcgi_param PATH_INFO $fastcgi_script_name; fastcgi_param REQUEST_METHOD $request_method; fastcgi_param QUERY_STRING $query_string; fastcgi_param CONTENT_TYPE $content_type; fastcgi_param CONTENT_LENGTH $content_length; fastcgi_pass_header Authorization; fastcgi_intercept_errors off; fastcgi_index index.html; index index.htm index.html; } error_page 500 502 503 504 /50x.html; location = /50x.html { root /etc/nginx/html/50x.html; } } # server { # listen 80; # server_name localhost; #charset koi8-r; #access_log logs/host.access.log main; # location / { # root html; # index index.html index.htm; # } #error_page 404 /404.html; # redirect server error pages to the static page /50x.html # #error_page 500 502 503 504 /50x.html; #location = /50x.html { root html; #} # proxy the PHP scripts to Apache listening on 127.0.0.1:80 # #location ~ \.php$ { # proxy_pass http://127.0.0.1; #} # pass the PHP scripts to FastCGI server listening on 127.0.0.1:9000 # #location ~ \.php$ { # root html; # fastcgi_pass 127.0.0.1:9000; # fastcgi_index index.php; # fastcgi_param SCRIPT_FILENAME /scripts$fastcgi_script_name; # include fastcgi_params; #} # deny access to .htaccess files, if Apache's document root # concurs with nginx's one # #location ~ /\.ht { # deny all; #} #} # another virtual host using mix of IP-, name-, and port-based configuration # #server { # listen 8000; # listen somename:8080; # server_name somename alias another.alias; # location / { # root html; # index index.html index.htm; # } #} # HTTPS server # #server { # listen 443; # server_name localhost; # ssl on; # ssl_certificate cert.pem; # ssl_certificate_key cert.key; # ssl_session_timeout 5m; # ssl_protocols SSLv2 SSLv3 TLSv1; # ssl_ciphers HIGH:!aNULL:!MD5; # ssl_prefer_server_ciphers on; # location / { # root html; # index index.html index.htm; # } #} } I can't quite tell if it's a server issue or a configuration issue: I've followed so many guides now I can't even count them all. The thing is that Django itself is working fine, and my permissions to the document root of the where the site files are stored is 777. Ontop of that, I have a git repo which works perfectly fine, and django, python, and runfcgi all start without issues. The same goes for gunicorn, when I do a gunicorn_django -b 0.0.0.0:8000 in my document root. Here is my output from that: 2012-04-15 05:17:37 [3124] [INFO] Starting gunicorn 0.14.2 2012-04-15 05:17:37 [3124] [INFO] Listening at: http://0.0.0.0:8081 (3124) 2012-04-15 05:17:37 [3124] [INFO] Using worker: sync 2012-04-15 05:17:37 [3127] [INFO] Booting worker with pid: 3127 As far as I know, everything seems fine, as well as error.log and access.log for nginx. The access log is completely blank, for that matter. I just feel lost here; what would be a step in the right direction to bebugging an issue such as this?

  Read the article

• Tips for maximizing Nginx requests/sec?

  - by linkedlinked

  I'm building an analytics package, and project requirements state that I need to support 1 billion hits per day. Yep, "billion". In other words, no less than 12,000 hits per second sustained, and preferably some room to burst. I know I'll need multiple servers for this, but I'm trying to get maximum performance out of each node before "throwing more hardware at it". Right now, I have the hits-tracking portion completed, and well optimized. I pretty much just save the requests straight into Redis (for later processing with Hadoop). The application is Python/Django with a gunicorn for the gateway. My 2GB Ubuntu 10.04 Rackspace server (not a production machine) can serve about 1200 static files per second (benchmarked using Apache AB against a single static asset). To compare, if I swap out the static file link with my tracking link, I still get about 600 requests per second -- I think this means my tracker is well optimized, because it's only a factor of 2 slower than serving static assets. However, when I benchmark with millions of hits, I notice a few things -- No disk usage -- this is expected, because I've turned off all Nginx logs, and my custom code doesn't do anything but save the request details into Redis. Non-constant memory usage -- Presumably due to Redis' memory managing, my memory usage will gradually climb up and then drop back down, but it's never once been my bottleneck. System load hovers around 2-4, the system is still responsive during even my heaviest benchmarks, and I can still manually view http://mysite.com/tracking/pixel with little visible delay while my (other) server performs 600 requests per second. If I run a short test, say 50,000 hits (takes about 2m), I get a steady, reliable 600 requests per second. If I run a longer test (tried up to 3.5m so far), my r/s degrades to about 250. My questions -- a. Does it look like I'm maxing out this server yet? Is 1,200/s static files nginx performance comparable to what others have experienced? b. Are there common nginx tunings for such high-volume applications? I have worker threads set to 64, and gunicorn worker threads set to 8, but tweaking these values doesn't seem to help or harm me much. c. Are there any linux-level settings that could be limiting my incoming connections? d. What could cause my performance to degrade to 250r/s on long-running tests? Again, the memory is not maxing out during these tests, and HDD use is nil. Thanks in advance, all :)

  Read the article

• Nginx error page using django master page

  - by user835199

  I am using python django to develop a web app and using nginx and gunicorn as servers. I need to define nginx error page (for error codes like 500, 501 etc), but i want to keep the layout same as that in other site pages. For site pages, i use the include functionality of django but in this case, since django won't preprocess the page, i need to create a pure html page. Is there a way to reuse the master page that i created in django for creating this error page?

  Read the article

• reliably restarting services using upstart or runit

  - by murtaza52

  I want to reliably restart my app and web server processes on crash. If I understand correctly, runit starts every service as a child process. If the child process crashes this sends a signal to the parent process which in turn respawns the service as a child. How does this work in the case of upstart. Does it also spawn a child process like runit? I am considering using runit for this. Is that needed, or is upstart good enough for this ? I am using nginx for my web server and gunicorn (python) for my app server.

  Read the article

• Enable basic auth sitewide and disabling it for subpages?

  - by piquadrat

  I have a relatively straight forward config: upstream appserver-1 { server unix:/var/www/example.com/app/tmp/gunicorn.sock fail_timeout=0; } server { listen 80; server_name example.com; location / { proxy_pass http://appserver-1; proxy_redirect off; proxy_set_header Host $host; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; auth_basic "Restricted"; auth_basic_user_file /path/to/htpasswd; } location /api/ { auth_basic off; } } The goal is to use basic auth on the whole website, except on the /api/ subtree. While it does work with respect to basic auth, other directives like proxy_pass are not in effect on /api/ as well. Is it possible to just disable basic auth while retaining the other directives without copy&pasting everything?

  Read the article

• email handling with inbox.py and nginx

  - by Matt Ball

  I have a Flask web application running behind gunicorn and Nginx. Nginx proxies any traffic to ivrhub.org to the correct flask app. I would very much like to use inbox.py to process some incoming email. Running inbox.py's example on my server and then sending an email to [email protected] does not work as I intended. The inbox.py server does not seem to receive anything but the email also does not bounce. I'm missing something conceptually -- is there a DNS setting I need to configure or something I need to adjust with Nginx?

  Read the article

• Sometimes my urls get masked with the IP address instead of the domain

  - by user64631

  I have a server with one A record that points to my IP address. I have nginx with gunicorn as a prefork which goes to my django application For most of my pages, the URL is always my domain name in the url bar. However if I go to mydomain.com/admin the url magically transforms into x.x.x.x/admin in the url bar of my browser. I thought that was weird but I ignored it figuring it only happened for admin so it wasnt that big of a deal. Then I installed django-registration. So when I go to mydomain.com/accounts/register the url is still mydomain.com/accounts/register in the url bar. but when I submit a form, the POST request goes to x.x.x.x/accounts/register which creates a cross domain error. So I decided that it wasnt isolated to the admin and I really need to fix what is going on. I have no idea what is going on and am completely lost.

  Read the article

• Forward Request to Multiple Servers

  - by cactuarz

  We have 2 servers. One is old server and another is the new one. Currently we about doing a migration because the old server is not capable enough to handle everyday requests. The specs are: Old server Ubuntu 10.04 Nginx as Reverse Proxy Apache WSGI Python/Django New Server Ubuntu 10.04 Nginx Gunicorn Python/Django Celery+Redis Our manager asked us to research if the old server can perform multiple forwarding to all incoming request, for example, set Nginx of old server to forward all request to both old and new server. The purpose is to perform unit testing to new server using old server as comparer, see if the new server is ready to take over the role. Please help, if there is an idea, or must install some engine, or what we do is impossible. Many thanks.

  Read the article

• What kind of hosting do I need?

  - by Robert Smith

  I migrated this question from serverfault. Hopefully this is the appropriate place. I have been trying to answer this question but I haven't found an specific answer to my situation. As I want to pay for what I need, I thought I could get a good answer here. I have a custom made forum (rather than a built-in forum like the ones you can find in plugins, e.g. WP-Forum or phpBB type of software) in Django. I don't want to use Apache and modwsgi because it's usually very memory-hungry and I can't afford a big server. I prefer a combination of nginx and gunicorn which I think is very efficient (maybe you can also tell me what you think about that). I'm expecting to receive 10,000 to 20,000 visits each month with 15,000 to 30,000 page impressions. I have reviewed some cloud services like Amazon EC2 or Rackspace and other more traditional services (Linodo). This site won't use videos or big images and I certainly don't need a huge amount of bandwidth (200GB would be definitely too much). I need shell access so shared hosting is out of the question. What do I need to run a website like that without problems? What about RAM? 256MB would be enough (that's the amount of RAM offered by small instances in Amazon and Rackspace)? Do you know of any alternative to those I mentioned? If you need more information to provide a useful answer, please don't hesitate to ask. By the way, I was told that Linodo is not all that different to Amazon EC2 but this website is supposed to work 24/7, so I can't take advantage of Linodo's flexibility regarding creating and deleting instances. Thanks in advance.

  Read the article

• PHP and Django: Nginx, FastCGI and Green Unicorn?

  - by littlejim84

  I'm curious... I'm looking to have a really efficient setup for my slice for a client. I'm not an expert with servers and so am looking for good solid resources to help me set this up... It's been recommended to me that using FastCGI for PHP, Green Unicorn (gunicorn) for Django and Nginx for media is a good combination to have PHP and Django running on the same slice/server. This is needed due to have a main Django website and admin, but also to have a PHP forum on there too. Could anyone push me to some useful resources that would help me set this up on my slice? Or at least, any views or comments on this particular setup?

  Read the article

• Django Asynchronous Processing

  - by freyrs

  Hello all, I have a bunch of Django requests which executes some mathematical computations ( written in C and executed via a Cython module ) which may take an indeterminate amount ( on the order of 1 second ) of time to execute. Also the requests don't need to access the database. Right now everything is synchronous ( using Gunicorn with sync worker types ) but I'd like to make this asynchronous and nonblocking. I am very new to asynchronous Django, and so my question is what is the best stack for doing this. Is this sort of process something a task queue is well suited for? Would anyone recommend Tornado + Celery + RabbitMQ, or perhaps something else? Thanks in advance!

  Read the article

• What's the most scalable way to handle somewhat large file uploads in a Python webapp?

  - by Jason Baker

  We have a web application that takes file uploads for some parts. The file uploads aren't terribly big (mostly word documents and such), but they're much larger than your typical web request and they tend to tie up our threaded servers (zope 2 servers running behind an Apache proxy). I'm mostly in the brainstorming phase right now and trying to figure out a general technique to use. Some ideas I have are: Using a python asynchronous server like tornado or diesel or gunicorn. Writing something in twisted to handle it. Just using nginx to handle the actual file uploads. It's surprisingly difficult to find information on which approach I should be taking. I'm sure there are plenty of details that would be needed to make an actual decision, but I'm more worried about figuring out how to make this decision than anything else. Can anyone give me some advice about how to proceed with this?

  Read the article

• What open source ecommerce webshops offer #1: usability, #2: PayPal integration, and #3: ease of administration and use

  - by Jonathan Hayward

  I've spent several days trying to deploy Satchmo, in the process asking several questions about deployment (http://stackoverflow.com/questions/11277407/can-anyone-explain-this-error-message-deploying-a-satchmo-project-under-gunicorn, http://stackoverflow.com/questions/11277685/is-there-a-howto-to-fcgi-for-deploying-satchmo, and http://stackoverflow.com/questions/11278295/what-is-the-most-stable-release-of-satchmo). Django's tagline is "The web framework for perfectionists with deadlines," and Satchmo's tagline is equally forceful: "The webshop for perfectionists with deadlines." I'm looking more to set up, configure, design, etc., rather than code for this one, and I'm taking a bit of a hint that for me at least the "with deadlines" bit is something that I cannot manage. Deployment has been a time sink. So, taking a step back, I don't specifically need to edit and extend the source code; what I want are first, good usability and a clean experience for the end-user, then being easy to deploy/install/manage/maintain, and enough so that even if you're having a slow day it should at most be one day's work to install, one day's work to get running, and one day's work to rebrand as white label (for simple branding). What ecommerce webshops should I be looking at?

  Read the article

• What's the best way to run Drupal and Django sites behind the same Varnish server?

  - by Alexis Bellido

  I have a high traffic website running with Drupal and Apache, five web servers behind a Varnish server load balancing. Let's say this site is example.com. I'm using five backends and a director like this in my default.vcl: director balancer round-robin { { .backend = web1; } { .backend = web2; } { .backend = web3; } { .backend = web4; } { .backend = web5; } } Now I'm working on a new Django project that will be a new section of this site running on example.com/new-section. After checking the documentation I found I can do something like this: sub vcl_recv { if (req.url ~ "^/new-section/") { set req.backend = newbackend; } else { set req.backend = default; } } That is, using a different backend for a subdirectory /new-section under the same domain. My question is, how do I make something like this work with my director and load balancing setup? I'm probably going to run two or more web servers (backends) with my new Django project, each one with a mix of Gunicorn, Nginx, and a few Python packages, and would like to put all of those in their own Varnish director to load balance. Is it possible to do use the above approach to decide which director to use?, like this: sub vcl_recv { if (req.url ~ "^/new-section/") { set req.director = newdirector; } else { set req.director = balancer; } } All suggestions welcome. Thanks!

  Read the article

• What kind of hosting do I need? [closed]

  - by Robert Smith

  I have been trying to answer this question but I haven't found an specific answer to my situation. As I want to pay for what I need, I thought I could get a good answer here. I have custom made forum (rather than a built-in forum like the ones you can find as plugins, e.g. WP-Forum or phpBB type of software) in Django. I don't want to use Apache and modwsgi because it's usually very memory-hungry and I can't afford a big server. I prefer a combination of nginx and gunicorn which I think is very efficient (maybe you can also tell me what you think about that). I'm expecting to receive 10,000 to 20,000 visits each month with 15,000 to 30,000 page impressions. I have reviewed some cloud services like Amazon EC2 or Rackspace and other more traditional services (Linodo). This site won't use videos or big images and I certainly don't need a huge amount of bandwidth (200GB would be definitely too much). I need shell access so shared hosting is out of the question. What do I need to run a website like that without problems? What about RAM? 256MB would be enough (that's the amount of RAM offered by small instances in Amazon and Rackspace)? Do you know of any alternative to those I mentioned? If you need more information to provide a useful answer, please don't hesitate to ask. Thanks a lot.

  Read the article

• 403 Forbidden serving static files from VirtualBox shared folder with nginx (Ubuntu 10.04LTS guest, Windows 7 host)

  - by Chris Pratt

  I'm working on a local development VM and trying to test serving my site with gunicorn and nginx as a reverse proxy for static resources only. The site loads minus static resources with user nginx; in nginx.conf. Attempting to load a static resource individually reveals a 403 Forbidden error. For background. The static resources are in a shared folder under /media/sf_work. All files are owned by root:vboxsf (VirtualBox default). My user account on the system has been added to the vboxsf group, and I have full access to the shared folder. For comparison, I tried changing the nginx.conf user to my user account. In that scenario, the static files did load, but then the homepage itself gives a 403 Forbidden error. So, I then tried adding the nginx user to the vboxsf group, but then everything gives a 403 Forbidden error. After further investigation it seems that if the nginx.conf user is in any group, it results in a 403 Forbidden. Any idea what could possibly be going on here?

  Read the article

• Python Django sites on Apache+mod_wsgi with nginx proxy: highly fluctuating performance

  - by Halfgaar

  I have an Ubuntu 10.04 box running several dozen Python Django sites using mod_wsgi (embedded mode; the faster mode, if properly configured). Performance highly fluctuates. Sometimes fast, sometimes several seconds delay. The smokeping graphs are al over the place. Recently, I also added an nginx proxy for the static content, in the hopes it would cure the highly fluctuating performance. But, even though it reduced the number of requests Apache has to process significantly, it didn't help with the main problem. When clicking around on websites while running htop, it can be seen that sometimes requests are almost instant, whereas sometimes it causes Apache to consume 100% CPU for a few seconds. I really don't understand where this fluctuation comes from. I have configured the mpm_worker for Apache like this: StartServers 1 MinSpareThreads 50 MaxSpareThreads 50 ThreadLimit 64 ThreadsPerChild 50 MaxClients 50 ServerLimit 1 MaxRequestsPerChild 0 MaxMemFree 2048 1 server with 50 threads, max 50 clients. Munin and apache2ctl -t both show a consistent presence of workers; they are not destroyed and created all the time. Yet, it behaves as such. This tells me that once a sub interpreter is created, it should remain in memory, yet it seems sites have to reload all the time. I also have a nginx+gunicorn box, which performs quite well. I would really like to know why Apache is so random. This is a virtual host config: <VirtualHost *:81> ServerAdmin [email protected] ServerName example.com DocumentRoot /srv/http/site/bla Alias /static/ /srv/http/site/static Alias /media/ /srv/http/site/media WSGIScriptAlias / /srv/http/site/passenger_wsgi.py <Directory /> AllowOverride None </Directory> <Directory /srv/http/site> Options -Indexes FollowSymLinks MultiViews AllowOverride None Order allow,deny allow from all </Directory> Ubuntu 10.04 Apache 2.2.14 mod_wsgi 2.8 nginx 0.7.65 Edit: I've put some code in the settings.py file of a site that writes the date to a tmp file whenever it's loaded. I can now see that the site is not randomly reloaded all the time, so Apache must be keeping it in memory. So, that's good, except it doesn't bring me closer to an answer... Edit: I just found an error that might also be related to this: File "/usr/lib/python2.6/subprocess.py", line 633, in __init__ errread, errwrite) File "/usr/lib/python2.6/subprocess.py", line 1049, in _execute_child self.pid = os.fork() OSError: [Errno 12] Cannot allocate memory The server has 600 of 2000 MB free, which should be plenty. Is there a limit that is set on Apache or WSGI somewhere?

  Read the article

< Previous Page | 1 2