ITB2017 - Nginx Effective High Availability Content Caching

NGINX, Inc. 2017
Using NGINX as an Effective
and Highly Available Content
Cache
Kevin Jones
Technical Solutions Architect
@webopsx

• Quick intro to…
• NGINX
• Content Caching
• Caching with NGINX
• How caching functionality works
• How to enable basic caching
• Advanced caching with NGINX
• How to increase availability using caching
• When and how to enable micro-caching
• How to ﬁne tune the cache
• How to architect for high availability
• Various conﬁguration tips and tricks!
• Various examples!
2
Agenda

MORE INFORMATION AT NGINX.COM
Solves complexity…
Web Server Load BalancerReverse Proxy Content Cache
Streaming
Media Server

350 million
total sites and counting… 
running on NGINX

53%of the Top 10,000 
most visited websites

36%of all instances on 
Amazon Web Services
Source: W3Techs December 2013 Web Server Survey

Contexts, directives and parameters… oh my.

9
user nginx;
worker_processes auto;
error_log /var/log/nginx/error.log notice;
pid /var/run/nginx.pid;
events {
worker_connections 1024;
}
http {
include /etc/nginx/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 /var/log/nginx/access.log main;
upstream api-backends {
server 10.0.1.11:8080;
server 10.0.1.12:8080;
}
server {
listen 10.0.1.10:80;
server_name example.com;
location / {
root /usr/share/nginx/html;
index index.html index.htm;
}
location ^~ /api {
proxy_pass http://api-backends;
}
}
include /path/to/more/virtual_servers/*.conf;
}
nginx.org/en/docs/dirindex.html
http context
server context
events context
main context
stream contextnot shown…
upstream context
location context

10
user nginx;
events {
}
http {
server 10.0.1.11:8080;
server 10.0.1.12:8080;
}
server {
listen 10.0.1.10:80;
location / {
}
location ^~ /api {
}
}
}
server directive
location directive
upstream directive
events directive
main directive

11
user nginx;
events {
}
http {
server 10.0.1.11:8080;
server 10.0.1.12:8080;
}
server {
listen 10.0.1.10:80;
location / {
}
location ^~ /api {
}
}
}
parameter
parameter
parameter
parameter

Source: W3Techs December 2013 Web Server Survey
Variables

13
user nginx;
events {
}
http {
server 10.0.1.11:8080;
server 10.0.1.12:8080;
}
server {
listen 10.0.1.10:80;
location / {
}
location ^~ /api {
}
}
}
nginx.org/en/docs/varindex.html
variables

14
http {
map $http_user_agent $dynamic {
“~*Mobile” mobile.example.com;
default desktop.example.com;
}
server {
listen 10.0.1.10:80;
location / {
}
location ^~ /api {
proxy_pass http://$dynamic;
}
}
}
variable
map (dynamic)

Server Name Identiﬁcation
(i.e. hostname routing)

16
http {
...
server {
listen 10.0.1.10:80;
server_name example.com *.website.com;
location / {
}
location ^~ /api {
}
}
}
SNI

18
http {
...
server {
if ( $blocked ) {
return 444;
}
listen 10.0.1.10:80;
server_name website.com *.example.com;
location / {
}
location ^~ /api {
}
}
}
Regex location matching
Conditional Routing

19
The Basics of Content Caching

20
Client
initiates request
(e.g. GET /ﬁle)
Proxy Cache
determines if response
is already cached if not
NGINX will fetch from the
origin server Origin Server
serves response
along with all
cache control headers
(e.g. Cache-Control,
Etag, etc..)
Proxy Cache
caches the response
and serves it to the client

21
Cache Headers
• Cache-Control - used to specify directives for caching mechanisms in both, requests and
responses. (e.g. Cache-Control: max-age=600 or Cache-Control: no-cache)
• Expires - contains the date/time after which the response is considered stale. If there is a Cache-
Control header with the "max-age" or "s-max-age" directive in the response, the Expires header is
ignored. (e.g. Expires: Wed, 21 Oct 2015 07:28:00 GMT)
• Last-Modified - contains the date and time at which the origin server believes the resource was last
modified. HTTP dates are always expressed in GMT, never in local time. Less accurate than the
ETag header (e.g. Last-Modified: Wed, 21 Oct 2015 07:28:00 GMT)
• ETag - is an identifier (or fingerprint) for a specific version of a resource. (e.g. ETag: “58efdcd0-268")

22
Content caching with NGINX is simple.

23
proxy_cache_path
proxy_cache_path path [levels=levels] [use_temp_path=on|off] keys_zone=name:size [inactive=time]
[max_size=size] [manager_files=number] [manager_sleep=time] [manager_threshold=time]
[loader_files=number] [loader_sleep=time] [loader_threshold=time] [purger=on|off] [purger_files=number]
[purger_sleep=time] [purger_threshold=time];
Syntax:
Default: -
Context: http
Documentation
http {
proxy_cache_path /tmp/nginx/micro_cache/ levels=1:2 keys_zone=large_cache:10m
max_size=300g inactive=14d;
...
}
Definition: Sets the path and other parameters of a cache. Cache data are stored in files. The file name in a cache is
a result of applying the MD5 function to the cache key.

24
proxy_cache_key
Documentation
server {
proxy_cache_key $scheme$proxy_host$request_uri$cookie_userid$http_user_agent;
...
}
proxy_cache_key string;Syntax:
Default: proxy_cache_key $scheme$proxy_host$request_uri;
Context: http, server, location
Deﬁnition: Deﬁnes a key for caching. Used in the proxy_cache_path directive.

25
proxy_cache
Documentation
location ^~ /api {
...
proxy_cache large_cache;
}
proxy_cache zone | off;Syntax:
Default: proxy_cache off;
Deﬁnition: Deﬁnes a shared memory zone used for caching. The same zone can be used in several places.

26
proxy_cache_valid
Documentation
location ~* .(jpg|png|gif|ico)$ {
...
proxy_cache_valid any 1d;
}
proxy_cache_valid [code ...] time;Syntax:
Default: -
Deﬁnition: Sets caching time for different response codes.

27
http {
proxy_cache_path /tmp/nginx/cache levels=1:2 keys_zone=cache:10m
max_size=100g inactive=7d use_temp_path=off;
...
server {
...
location / {
...
proxy_pass http://backend.com;
}
location ^~ /images {
...
proxy_cache cache;
proxy_cache_valid 200 301 302 12h;
proxy_pass http://images.origin.com;
}
}
}
Basic Caching

28
Client
NGINX Cache
Origin Server
Cache Memory Zone
(Shared across workers)
1. HTTP Request:
GET /images/hawaii.jpg
Cache Key: http://origin/images/hawaii.jpg
md5 hash: 51b740d1ab03f287d46da45202c84945
2. NGINX checks if hash exists in memory. If it does
not the request is passed to the origin server.
3. Origin server
responds
4. NGINX caches the response to disk
and places the hash in memory
5. Response is served to client

29
NGINX Processes
# ps aux | grep nginx
root 14559 0.0 0.1 53308 3360 ? Ss Apr12 0:00 nginx: master process /usr/
sbin/nginx -c /etc/nginx/nginx.conf
nginx 27880 0.0 0.1 53692 2724 ? S 00:06 0:00 nginx: worker process
nginx 27881 0.0 0.1 53692 2724 ? S 00:06 0:00 nginx: worker process
nginx 27882 0.0 0.1 53472 2876 ? S 00:06 0:00 nginx: cache manager process
nginx 27883 0.0 0.1 53472 2552 ? S 00:06 0:00 nginx: cache loader process
• Cache Manager - activated periodically to check the state of the cache. If the cache
size exceeds the limit set by the max_size parameter to the proxy_cache_path directive,
the cache manager removes the data that was accessed least recently
• Cache Loader - runs only once, right after NGINX starts. It loads metadata about
previously cached data into the shared memory zone.

30
Caching is not just for HTTP
HTTP
FastCGI
UWSGI
SCGI
Memcache
Tip: NGINX can also be used to cache other backends using their unique cache directives. (e.g. fastcgi_cache,
uwsgi_cache and scgi_cache)
Alternatively, NGINX can also be used to retrieve content directly from a memcached server.

31
Initial… Tips and Tricks!

32
log_format main 'rid="$request_id" pck="$scheme://$proxy_host$request_uri" '
'ucs="$upstream_cache_status" '
'site="$server_name" server="$host” dest_port="$server_port" '
'dest_ip="$server_addr" src="$remote_addr" src_ip="$realip_remote_addr" '
'user="$remote_user" time_local="$time_local" protocol="$server_protocol" '
'status="$status" bytes_out="$bytes_sent" '
'bytes_in="$upstream_bytes_received" http_referer="$http_referer" '
'http_user_agent="$http_user_agent" nginx_version="$nginx_version" '
'http_x_forwarded_for="$http_x_forwarded_for" '
'http_x_header="$http_x_header" uri_query="$query_string" uri_path="$uri" '
'http_method="$request_method" response_time="$upstream_response_time" '
'cookie="$http_cookie" request_time="$request_time" ';
Logging is your friend…
Tip: The more relevant information in your log the better. When troubleshooting you can easily add the proxy
cache KEY to the log_format for debugging. For a list of all variables see the “Alphabetical index of
variables” on nginx.org.

33
server {
...
# add HTTP response headers
add_header CC-X-Request-ID $request_id;
add_header X-Cache-Status $upstream_cache_status;
}
Adding response headers…
Tip: Using the add_header directive you can add useful HTTP response headers allowing you to debug
your NGINX deployment rather easily.

34
Cache Status
• MISS – The response was not found in the cache and so was fetched from an origin server. The response
might then have been cached.
• BYPASS – The response was fetched from the origin server instead of served from the cache because the
request matched a proxy_cache_bypass directive. The response might then have been cached.
• EXPIRED – The entry in the cache has expired. The response contains fresh content from the origin
server.
• STALE – The content is stale because the origin server is not responding correctly, and
proxy_cache_use_stale was configured.
• UPDATING – The content is stale because the entry is currently being updated in response to a previous
request, and proxy_cache_use_stale updating is configured.
• REVALIDATED – The proxy_cache_revalidate directive was enabled and NGINX verified that the current
cached content was still valid (ETag, If‑Modified‑Since or If‑None‑Match).
• HIT – The response contains valid, fresh content direct from the cache.

35
# curl -I 127.0.0.1/images/hawaii.jpg
HTTP/1.1 200 OK
Server: nginx/1.11.10
Date: Wed, 19 Apr 2017 22:20:53 GMT
Content-Type: image/jpeg
Content-Length: 21542868
Connection: keep-alive
Last-Modified: Thu, 13 Apr 2017 20:55:07 GMT
ETag: "58efe5ab-148b7d4"
OS-X-Request-ID: 1e7ae2cf83732e8859bc3e38df912ed1
CC-X-Request-ID: d4a5f7a8d25544b1409c351a22f42960
X-Cache-Status: HIT
Accept-Ranges: bytes
Using cURL to Debug…
Tip: Use cURL or Chrome developer tools to grab the request ID or other various headers useful for
debugging.

36
# grep -ri d4a5f7a8d25544b1409c351a22f42960 /var/log/nginx/adv_access.log
rid="d4a5f7a8d25544b1409c351a22f42960" pck="http://origin/images/hawaii.jpg"
site="webopsx.com" server="localhost” dest_port="80" dest_ip=“127.0.0.1" ...
# echo -n "http://origin/images/hawaii.jpg" | md5sum
51b740d1ab03f287d46da45202c84945 -
# tree /tmp/nginx/micro_cache/5/94/
/tmp/nginx/micro_cache/5/94/
!"" 51b740d1ab03f287d46da45202c84945
0 directories, 1 file
Troubleshooting the Proxy Cache
Tip: A quick and easy way to determine the hash of your cache key can be accomplished using echo, pipe and
md5sum

37
# head -n 14 /tmp/nginx/micro_cache/5/94/51b740d1ab03f287d46da45202c84945
??X?X??Xb?!bv?"58efe5ab-148b7d4"
KEY: http://origin/images/hawaii.jpg
HTTP/1.1 200 OK
Server: nginx/1.11.10
Date: Wed, 19 Apr 2017 23:51:38 GMT
Content-Type: image/jpeg
Content-Length: 21542868
Last-Modified: Thu, 13 Apr 2017 20:55:07 GMT
Connection: keep-alive
ETag: "58efe5ab-148b7d4"
OS-X-Request-ID: 1e7ae2cf83732e8859bc3e38df912ed1
Accept-Ranges: bytes
?wExifII>(i?Nl?0230??HH??
Cache Contents

38
Micro-Caching
“Size matters not.”

39
Static Content
• Images
• CSS
• Simple HTML
User Content
• Shopping Cart
• Unique Data
• Account Data
Dynamic Content
• Blog Posts
• Status
• API Data (Maybe?)
Easy to cache Cannot CacheMicro-cacheable!
Types of Content
Documentation

40
http {
upstream backend {
keepalive 20;
server 127.0.0.1:8080;
}
proxy_cache_path /var/nginx/micro_cache levels=1:2 keys_zone=micro_cache:10m
max_size=100m inactive=600s;
...
server {
listen 80;
...
proxy_cache micro_cache;
proxy_cache_valid any 1s;
location / {
proxy_http_version 1.1;
proxy_set_header Connection "";
proxy_set_header Accept-Encoding "";
proxy_pass http://backend;
}
}
}
Enable keepalives on upstream
Set proxy_cache_valid to any
status with a 1 second value
Set required HTTP version and
pass HTTP headers for keepalives
Set short inactive parameter

41
proxy_cache_lock
Documentation
proxy_cache_lock on | off;Syntax:
Default: proxy_cache_lock off;
Deﬁnition: When enabled, only one request at a time will be allowed to populate a new cache element identiﬁed
according to the proxy_cache_key directive by passing a request to a proxied server.
Other requests of the same cache element will either wait for a response to appear in the cache or the
cache lock for this element to be released, up to the time set by the proxy_cache_lock_timeout directive.
Related: See the following for tuning…
• proxy_cache_lock_age,
• proxy_cache_lock_timeout

42
proxy_cache_use_stale
Documentation
location /contact-us {
...
proxy_cache_use_stale error timeout updating http_500 http_502 http_503 http_504;
}
proxy_cache_use_stale error | timeout | invalid_header | updating | http_500 | http_502 | http_503 |
http_504 | http_403 | http_404 | http_429 | off ...;
Syntax:
Default: proxy_cache_use_stale off;
Deﬁnition: Determines in which cases a stale cached response can be used during communication with the proxied
server.

43
http {
upstream backend {
keepalive 20;
server 127.0.0.1:8080;
}
proxy_cache_path /var/nginx/micro_cache levels=1:2 keys_zone=micro_cache:10m
max_size=100m inactive=600s;
...
server {
listen 80;
...
proxy_cache micro_cache;
proxy_cache_valid any 1s;
proxy_cache_lock on;
proxy_cache_use_stale updating;
location / {
...
proxy_set_header Connection "";
proxy_set_header Accept-Encoding "";
proxy_pass http://backend;
}
}
}
Final optimization

44
Further Tuning and Optimization

45
proxy_cache_revalidate
Documentation
proxy_cache_revalidate on | off;Syntax:
Default: proxy_cache_revalidate off;
Definition: Enables revalidation of expired cache items using conditional GET requests with the “If-Modified-Since”
and “If-None-Match” header fields.

46
proxy_cache_min_uses
Documentation
location ~* /legacy {
...
proxy_cache_min_uses 5;
}
proxy_cache_min_uses number;Syntax:
Default: proxy_cache_min_uses 1;
Deﬁnition: Sets the number of requests after which the response will be cached. This will help with disk utilization and
hit ratio of your cache.

47
proxy_cache_methods
Documentation
location ~* /data {
...
proxy_cache_methods GET HEAD POST;
}
proxy_cache_methods GET | HEAD | POST …;Syntax:
Default: proxy_cache_methods GET HEAD;
Deﬁnition: NGINX only caches GET and HEAD request methods by default. Using this directive you can add
additional methods.
If you plan to add additional methods consider updating the cache key to include the $request_method
variable if the response will be different depending on the request method.

48
proxy_buffering
Documentation
proxy_buffering on | off;Syntax:
Default: proxy_buffering on;
Definition: Enables or disables buffering of responses from the proxied server.
When buffering is enabled, nginx receives a response from the proxied server as soon as possible, saving
it into the buffers set by the proxy_buffer_size and proxy_buffers directives. If the whole response does not
fit into memory, a part of it can be saved to a temporary file on the disk.
When buffering is disabled, the response is passed to a client synchronously, immediately as it is received.

49
location ^~ /wordpress {
...
proxy_cache cache;
proxy_ignore_headers Cache-Control;
}
Override Cache-Control headers
Tip: By default NGINX will honor all Cache-Control headers from the origin server, in turn not caching
responses with Cache-Control set to Private, No-Cache, No-Store or with Set-Cookie in the response
header.
Using proxy_ignore_headers you can disable processing of certain response header ﬁelds from the
proxied server.

50
location / {
...
proxy_cache cache;
proxy_cache_bypass $cookie_nocache $arg_nocache $http_cache_bypass;
}
Can I Punch Through the Cache?
Tip: If you want to disregard the cache and go strait to the origin for a response, you can use the
proxy_cache_bypass directive.

51
proxy_cache_purge
Documentation
proxy_cache_methods string ...;Syntax:
Default: -
Deﬁnition: Deﬁnes conditions under which the request will be considered a cache purge request. If at least one value
of the string parameters is not empty and is not equal to “0” then the cache entry with a corresponding
cache key is removed.
The result of successful operation is indicated by returning the 204 (No Content) response.
Note: NGINX Plus only feature

52
proxy_cache_path /tmp/cache keys_zone=mycache:10m levels=1:2 inactive=60s;
map $request_method $purge_method {
PURGE 1;
default 0;
}
server {
listen 80;
server_name www.example.com;
location / {
proxy_pass http://localhost:8002;
proxy_cache mycache;
proxy_cache_purge $purge_method;
}
}
Example Cache Purge Conﬁguration
Tip: Using NGINX Plus, you can issue unique request methods to invalidate the cache

54
http {
proxy_cache_path /path/to/hdd1 levels=1:2 keys_zone=my_cache_hdd1:10m
max_size=10g inactive=60m use_temp_path=off;
proxy_cache_path /path/to/hdd2 levels=1:2 keys_zone=my_cache_hdd2:10m
max_size=10g inactive=60m use_temp_path=off;
split_clients $request_uri $my_cache {
50% “my_cache_hdd1”;
50% “my_cache_hdd2”;
}
server {
...
location / {
proxy_cache $my_cache;
proxy_pass http://my_upstream;
}
}
}
Split the Cache Across HDDs

55
Using NGINX for Byte Range Caching

56
http {
proxy_cache_path /tmp/mycache keys_zone=mycache:10m;
server {
listen 80;
proxy_cache mycache;
slice 1m;
proxy_cache_key $host$uri$is_args$args$slice_range;
proxy_set_header Range $slice_range;
proxy_cache_valid 200 206 1h;
location / {
proxy_pass http://origin.example.com;
}
}
Split the Cache Across HDDs
Tip: Using the split_client directive, NGINX will perform a hash function on a variable of your choice and
based on that hash will dynamically set a new variable that can be used elsewhere in the conﬁguration.

57
Architecting for High Availability

58
Two Approaches
• Sharded (High Capacity)
• Shared (Replicated)

59
Shared Cache Clustering
Tip: If your primary goal is to achieve high availability while minimizing load on the origin servers, this scenario
provides a highly available shared cache.

60
And Failover…
Tip: In the event of a failover there is no loss in cache and the origin does not suffer unneeded proxy requests.

61
Sharding your Cache
Tip: If your primary goal is to create a very high‑capacity cache, shard (partition) your cache across multiple
servers. This in turn maximizes the resources you have while minimizing impact on your origin servers
depending on the amount of cache servers in your cache tier.

62
upstream cache_servers {
hash $scheme$proxy_host$request_uri consistent;
server cache1.example.com;
}
Hash Load Balancing
Tip: Using the hash load balancing algorithm, we can specify the proxy cache key. This allows each resource to
be cached on only one backend server.

63
Combined Load Balancer and Cache
Tip: Alternatively, It is possible to consolidate the load balancer and cache tier into one with the use of a
various NGINX directives and parameters.

64
Multi-Tier with “Hot Cache”
Tip: If needed, a “Hot Cache Tier” can be enabled on the load balancer layer which will give you the same high
capacity cache and provide a high availability of speciﬁc cached resources.

Documentation
• https://nginx.org
• https://nginx.com
Blog
• https://www.nginx.com/blog/nginx-caching-guide/
• https://www.nginx.com/blog/beneﬁts-of-microcaching-nginx/
• https://www.nginx.com/blog/shared-caches-nginx-plus-cache-clusters-part-1/
• https://www.nginx.com/blog/shared-caches-nginx-plus-cache-clusters-part-2/
• https://www.nginx.com/blog/smart-efﬁcient-byte-range-caching-nginx/
Webinar
• https://www.nginx.com/resources/webinars/content-caching-nginx-plus/
65
Links

Thank You
66
https://www.nginx.com/blog/author/kjones/
@webopsx
Kevin Jones
Technical Solutions Architect
NGINX Inc.
kevin@nginx.com
https://www.slideshare.net/KevinJones62
https://www.linkedin.com/in/kevin-jones-19b17b47/

ITB2017 - Nginx Effective High Availability Content Caching

Related slideshows

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

Recommended for you

More Related Content

What's hot

What's hot (20)

Similar to ITB2017 - Nginx Effective High Availability Content Caching

Similar to ITB2017 - Nginx Effective High Availability Content Caching (20)

More from Ortus Solutions, Corp

More from Ortus Solutions, Corp (20)

Recently uploaded

Recently uploaded (20)

ITB2017 - Nginx Effective High Availability Content Caching