CACHINGMEMCACHED vs. REDIS vs. FILESYSTEM
Brown Bag Session
Rafi AdnanFaisal Islam
What is caching in web application?
Web application caching is the process of storing dynamically generated data for reuse and leaving data closer to the end user.
How Caching Works
Traditional Data Query (Without Cache)
Query
Get Rows From Database
Return Rows to Application
Cached Data QueryQuery
Data In
Cache
Get Rows From Database
Return Row to Application
Insert Rows Into
Cache
Cache StorageDatabase
Yes
No
Why we use?➔ Better performance
◆ System loads faster/better responsiveness
➔ Better scalability
◆ Limit bottlenecks in a system
➔ Better robustness
◆ Can support more load
When to use?
Profile your application to find operations that consume significant CPU time and/or memory
try - “New Relic”
Don’t be premature
CachingTerminology
[ Cache Hit ]when requested data is contained in the cache
[ Cache Miss ]when requested not in the cached, has to be recomputed or fetched from original storage
[ Cache Key ]unique identifier for a data item in the cache
[ Expiration ]item expires at a specific date (absolute), specifies how long after an item was last accessed that is expires (sliding)
[ Backing store ]persist cached data on disk
[ Cache Scavenging ]deleting items from the cache when memory is scarce
[ Local Cache ]caching data on clients rather than on servers
[ Distributed Cache ]extension of the traditional concept of cache that may span multiple servers
Advantages❖ Speed up application
❖ Less Resources Used
❖ Reuse content
Disadvantages
❖ Stale Data
❖ Overhead
❖ Complexity
Type Of Caching❑ Client Caching - browser caches URLs for future uses - Mozilla Firefox, Google Chrome
❑ Proxy Caching- proxy server caches most requested URLs
- Varnish ( reverse proxy ) - CDN ( forward proxy )
❑ Server-side Caching- server side cache reduces load on server
- File System Cache - In-Memory Cache (MemCached, Redis)
File Cache
A file of data on a local hard drive. When downloaded data are temporarily stored on the
user's local disk or on a local network disk, it speeds up retrieval the next time the user wants that same data (Web page, graphic, etc.) from the Internet or
other remote source.
In-Memory means We are Bound By RAM
MemCached
Memcached is an in-memory key-value store for small chunks of arbitrary data
(strings, objects) from results of database calls, API calls, or page
rendering.
Redis
Redis is an open source, advanced key-value store. It is often referred to as a "data structure server" since keys can contain strings, hashes, lists, sets and
sorted sets.
How to use?
[ File System Caching ]
➔ It works without any additional server. Most of the CMS by default use file cache without any configurations.
[ In-Memory Caching ]
➔ Need to install and configure cache server
➔ Need client library to access cache
➔ Most of the popular framework has very good built-in or third-party library.
In-Memory Cache ( Server - Client )
Installing memcached server (ubuntu):
sudo apt-get install memcached
Installing memcached client(ubuntu):
PHP :
sudo apt-get install php5-memcached
Ruby on Rails :
Popular gem called Dalli (https://github.com/mperham/dalli)
Installing redis server (ubuntu):
sudo apt-get install redis-server
Redis client 3rd-party packages:
PHP : Predis , phpredis
Ruby on Rails : redis-rb, em-hiredis, redic
Python : redis-py, txredisapi, brukva
(http://redis.io/clients)
MemCached
➔ Dead Simple & Battle Tested
➔ Fast
➔ Non-Blocking get()/set()
➔ Multi-Threaded
➔ Consistent Hashing
Code Example - Memcached$memCached = new Memcached();
//connect with memcached server
$memCached->addServer('127.0.0.1', '11211');
const INT_EXPIRATION_TIME_IN_SECONDS = 2;
//Set cache
$memCached->set('key_1', 'serialize data',
INT_EXPIRATION_TIME_IN_SECONDS);
//get cache
//if call unique_key after 2 seconds it will return
false
$uniqueKeyValue = $memCached->get('key_1');
var_export($uniqueKeyValue);
Output :
'serialize data'
Client
Port
Server Addr
Code Example - Memcached$memCached = new Memcached();//connect with memcached server$memCached->addServer('127.0.0.1', '11211');//consider our data set below$employeeId = 1000;$employeeData = array( 'name' => 'Rafi Adnan', 'designation' => 'Software Engineer', 'joining_date' => '2014-09-15');$memCached->set($employeeId, json_encode($employeeData));// Returns employee details as json$employeeDetail = $memCached->get($employeeId);var_export($employeeDetail);
Output :
{"name":"Rafi Adnan","designation":"Software Engineer","joining_date":"2014-09-15"}
What if I don't want all the data?
➔ What if I just want the name?
➔ 64 bytes for the object vs 16 bytes for just the name?
➔ 4X network traffic
➔ More work for the application
Redis: Data Types
➔ Strings ( just like Memcached )
➔ Lists
➔ Sets
➔ Sorted Sets
➔ Hashes
Redis : Lists
➔ Stored in sorted order
➔ Can push/pop
➔ Fast head/tail access
➔ Index access
Code Example - Redis : Lists
$redisClient = new Predis\Client();
$redisClient->lpush('employees', 'Faisal');$redisClient->lpush('employees', 'Rafi');
$arrEmployeeList = $redisClient->lrange('employees', 0, -1);var_export($arrEmployeeList);
$redisClient->rpush('employees', 'Akhtar');
$arrEmployeeList = $redisClient->lrange('employees', 0, -1);var_export($arrEmployeeList);
array ( 0 => 'Rafi', 1 => 'Faisal',)
array ( 0 => 'Rafi', 1 => 'Faisal', 2 => 'Akhtar',)
Redis : Sets
➔ Unordered collections of strings
➔ Unique ( no repeated members )
➔ diff, intersect, merge
Code Example - Redis : Sets$redisClient = new Predis\Client();
$redisClient->sadd('employees', 'Rafi Adnan');$redisClient->sadd('employees', 'Saeed Ahmed');$redisClient->sadd('employees', 'Faisal Islam');
$redisClient->sadd('formerEmployees', 'Saeed Ahmed');
$currentEmployees = $redisClient->sdiff('employees', 'formerEmployees');
var_export($currentEmployees); array ( 0 => 'Rafi Adnan', 1 => 'Faisal Islam',)
Code Example - Redis : Hashes
$redisClient = new Predis\Client();
$redisClient->hset('employees', 'teamCount', 10);$redisClient->hset('employees', 'rubyTeam', 9);$redisClient->hset('employees', 'phpTeam', 1);
$teamCount = $redisClient->hget('employees', 'teamCount');var_export($teamCount);
$teamStatus = $redisClient->hgetall('employees');var_export($teamStatus);
array ( 'teamCount' => '10', 'rubyTeam' => '9', 'phpTeam' => '1',)
‘10’ - string
Redis: The Bad
➔ Single-Threaded
➔ Limited client support for consistent hashing
➔ Significant overhead for persistence
➔ Not widely deployed
Memcached vs. RedisMemcached Redis
(multi) get ✔ ✔
(multi) set ✔ ✔
increment/decrement ✔ ✔
delete ✔ ✔
expiration ✔ ✔
range queries ✔
data types ✔
persistence - ✔
multi-threaded ✔
replication - ✔
Die;
Questions, comments...
