1. Scale Your Data Tier with
Windows Server AppFabric
Chris Dufour, ASP MVP
Software Architect, Compuware
chris.dufour@wigets.net

2. Agenda
The Problem
A Possible Solution
New Possibilities
What to do next

3. Simple
Do the simplest thing
that will possibly work

4. Scalability
Able to support the
required quality of
service as the system
load increases
-Wiktionary

5. Typical Web Architecture
•Need to get routed to
same machine (i.e. sticky
sessions)
Users
•Each machine round trips
for data
•Data that is cached is
stored in the memory of
one server
Web
Tier
•CPU and disk can get
saturated due to traffic
•Service access is slow
Data
Tier

6. Web Explosion
Database
IIS/ASP.NET
Application
Web
Site’s too
slow!!
Database is
hot!!
Where did
my
shopping
cart go?
IIS/ASP.NET
Application
IIS/ASP.NET
Application
Servers are
crashing
Services are
slow

7. Agenda
The Problem
A Possible Solution
New Possibilities
What to do next

8. Data Near Processing
Database
Smart Client Browser
Cache
CacheCache
CacheCache

9. Good but…
 Cache is scoped to machine / process
◦ Machines die
◦ Processes recycle
 Cache memory is limited

10. What if?
 You could have as much cache as you
wanted?
 You could share a giant cache across
servers, services and even clients?
 What if this was something you could
simply add to the platform for
1
free?
1Some features may require
certain editions of Windows
Server

11. Windows Server AppFabric
AppFabric
CACHING
MONITORING
WORKFLOW HOSTING
SERVICE HOSTING
SCALE OUTHIGH AVAILABILITY MANAGEMENT

12. Unified Cache View
What is AppFabric Caching?
 An explicit, distributed, in-memory
application cache for all kinds of data
Caching clients
can be across
machines or
processes
Clients Access
the Cache as if it
was a large
single cache
Cache Layer
distributes data
across the
various cache
nodes

13. AppFabric Cache
•No need for sticky sessionsUsers
•Cached data available to all
servers
Web
Tier
•Multiple machines means
scale and high-availability
•Lots of cache memory
Caching
Tier
•Reduces load on database
•Duplicate Service Calls
eliminated
Data
Tier

14. Data Distribution - Partitioned Cache
… Web Tier
Cache
Tier
CacheService
CacheService
Cache
Service
A C EDB G IHF
ASP.Net App
Caching Client
ASP.Net App
Caching Client
ASP.Net App
Caching Client
AC ED BGI HF
 Scale on Data Size
◦ More machines => More memory to cache
 Scale on Cache Throughput
◦ More machines => keys distributed across more machines => better
throughput

15. Scale Test Output
Load
Throughput
Latency
1 Cache Server
As load increases,
throughput fails
to scale
latency increases
Caching
Tier

16. Add a Second Cache Server
Load
Throughput
Latency
Load Max
Throughput
increases
Latency decreases
Caching
Tier

17. Add a Second Cache Server
Load
Throughput
Latency
Caching
Tier

18. Associated Press
 Caches metadata
and news
 Serves 16 million
hits per day
 Increased the
amount of cached
data 6 times.

19. System.Web.Cache

20. AppFabric DataCache

21. Caching
Service
Caching
Service
Cachin
g
Service
Configuration Store
Install
AppFabric
Copy Client DLLs
…Application Application Application<hosts>
<host name="BL1CDB8083714“
cachePort="22233"
cacheHostName="DistributedCacheService"/>
…..
</hosts>
<localCache isEnabled=“true" ../>
<security … />
Caching Access
Layer
Caching Access
Layer
Caching Access
Layer
.NET 3.5 SP1 OR .NET 4
Configure
AppFabric
.NET 4
Update Web.config
Deployment

22. Usage Pattern – Cache Aside
(Explicit Caching)
// Read from Cache
Toy toyObj = (Toy)
catalog.Get("toy-101");
Application
Caching
Service
Database
// If Not present in the cache
if (toyObj == null)
{
// Read from backend..
toyObj = ReadFromDatabase();
// Populate Cache
catalog.Put("toy-101", toyObj);
return toyObj;
}
Caching Access
Layer

23. Administration
 PowerShell cmdlets are used to administer
the cache cluster
 Rich set of cmdlets for
◦ Cache cluster management
◦ Cache creation and monitoring

24. Hello AppFabric Cache

25. Using PowerShell
Remember – PowerShell can also
be called from .NET Code!

26. Security
 Domain Based Security Option
◦ Domain Account / Local Account based
Authentication
◦ Only authorized servers can join the cluster
◦ Only authorized clients can connect to the
cluster
 Transport Level Security
◦ Turn on/off Signing or Encryption
 Can turn off Cache Security
◦ Use Firewalls, IPSec, VLANs to protect cache
grant-cacheallowedclientaccount RedDomainMachine1$
grant-cacheallowedclientaccount RedDomainJohn

27. Logical Hierarchy
 Machine
 Host
◦ Physical processes hosting
AppFabric Caching instance.
 Named Caches
◦ Can span across machines
◦ Defined in the configuration file
 Regions
◦ Physically co-located Container of
Cache Items
◦ May be implicit or explicitly created
 Cache Item
◦ Key, Payload (Object ), Tags, TTL,
Timestamps, Version
Regions
Region A
Key Payload Tags
121 xxxx “Toy” “Child”
123 yyyy “Toy” “Chair”..
AppFabric
Caching
Service
Named Cache : Product Catalog
Named Cache : Electronics Inventory
AppFabric
Caching
Service
AppFabric
Caching
Service
AppFabric
Caching
Service

28. AppFabric Caching API
// Create instance of cachefactory (reads appconfig)
DataCacheFactory fac = new DataCacheFactory();
// Get a named cache from the factory
DataCache catalog = fac.GetCache("catalogcache");
// Simple Get/Put
catalog.Put("toy-101", new Toy("Puzzle", .,.));
// From the same or a different client
Toy toyObj = (Toy)catalog.Get("toy-101");
// Region based Get/Put
catalog.CreateRegion("toyRegion");
// Both toy and toyparts are put in the same region
catalog.Put("toy-101", new Toy( .,.), “toyRegion”);
Catalog.Put("toypart-100", new ToyParts(…), “toyRegion”);
Toy toyObj = (Toy)catalog.Get("toy-101“,"toyRegion");

29. Access APIs – Tagging Items
Tag hotItem = new Tag("hotItem");
catalog.Put("toy-101", new Toy("Puzzle"),
new Tag[]{hotItem}, “toyRegion”);
catalog.Put("toy-102", new Toy("Bridge"), “toyRegion”);
// From the same or a different client
List<KeyValuePair<string, object>> toys =
catalog.GetAnyMatchingTag("toyRegion", hotItem);

30. Types of Data
Reference Activity Resource
Primary Read Only Read-Write Not
shared
Read-Write, Shared
Catalog Data Shopping Cart Auction Data/Seat
Assignment
Web Tier
Distributed Cache
Shopping Cart
Grocery
Catalog
Grocery
Inventory
Grocery Shop

31. Application
Application
AppFabric Caching Client
Reference Data – Performance
 Catalog data doesn’t change often
 Unnecessary network cost to access from different machines
 Solution – Local Cache
Put(K2, v3)
Routing Table
Cache2Cache1
Primary for K2,V2
K2, V2
Primary for K1,V1
K1, V1
Cache3
Primary for K3,V3
K3, V3
AppFabric Caching Client
Local Cache
Routing Table
K2,
V2
Get(K2)
Get(K2)
K2, V3

32. Reference Data – Bulk Get
 Bulk Fetch from region
◦ 200-300k ops per second
◦ Fewer network calls
Catalog.BulkGet(
new List<string>(){“toy-101”, “toy-102”} ,
“toyRegion”);

33. Activity Data – Session Integration
…
Session State
stored in
AppFabric
Caching
CacheService
Caching
Service
Load Balance Requests
No more sticky routing
CachingService
Scale your
Session Store
Dynamically
Highly
Available
Drop in AppFabric
Caching
SessionStoreProvider
Allows session state to be shared amongst multiple applications
Application
Caching Access
Layer
Application Application
Caching Access
Layer
Caching Access
Layer
<sessionState mode="Custom“
customProvider="SessionStoreProvider">
<providers>
<add name="SessionStoreProvider"
type=“Microsoft.Data.Caching.DataCacheSessionStoreProvider,
ClientLibrary“
cacheName="<YourNamedCache>"/>
</providers>
</sessionState>

34. Application
Application
(K2, V2)
Cache2Cache1 Cache3
Primary for
Activity Data - Availability
Get(K2)
Primary for Primary for
K3, V3
AppFabric Caching Client
Routing Table
K2,V2
PUT
Secondary for
K2, V2
K1, V1
Secondary for
K3, V3
Secondary for
K1, V1
AppFabric Caching Client
Routing Table
K2, V2
Replicatio
n AgentK2, V2

35. Resource Data - Optimistic Locking
 GetCacheItem returns a version object
 Every update to an object internally increments it's version
 Supply the version obtained along with the Put/Remove
 Put/Remove will succeed only if the passed in version matches
the version in the cache
Version Based Update
Time Client1 Client2 (Different Thread or
process)
T0 CacheItem item =
catalog.GetCacheItem(“PlayerRe
gion”, ”Zune”);
CacheItem item =
catalog.GetCacheItem(“PlayerRe
gion”,
”Zune”);
T1 ((ZuneObject)item.Object).inven
tory --;
((ZuneObject)item.Object).invent
ory--;
T2 catalog.Put(“PlayerRegion”,
“Zune”,
item.Object,
item.Version);
T3 catalog.Put(“PlayerRegion”,
“Zune”,
item.Object,
item.Version);
// Version mismatch
// Client must retry again
Two clients access the same
item
Both update the item
Second Client gets in first;
put succeeds because item
version matches; atomically
increments the version
First client tries put;
Fails because the versions
don’t match

36. K1
Resource Data - Pessimistic Locking
 Take locks on non-existent keys
◦ Allows you to co-ordinate calls for data
Client1:
GetAndLock ("k1")
Client2:
GetAndLock ("k1")
Client3:
Get ("k1")
Regular Get
succeeds
GetAndLock gets lock
handle
Other GetAndLock on
same item fails

37. Data Race
GET GETGET

38. Lock Non-Existent Key
GET/LOCK GET/LOCKGET/LOCK
Cache
Service
Caching
Service
Cache
Service

39. Composite Race
CALL WAITWAIT
Cache
Service
Caching
Service
Cache
Service

40. Composite Race
PUT
UNLOCK
GETGET
Cache
Service
Caching
Service
Cache
Service

41. Resource/Activity Data
– Tracking Changes
 Cache Event notifications
 Register on any client to notify changes
 Batched Notifications
DataCache.RegisterCacheLevelCallback(
int filter, DataCacheChangeCallback delegate);
DataCache.RegisterRegionLevelCallback(
String region, int filter, DataCacheChangeCallback delegate);
DataCache.RegisterKeyLevelCallback(
String region, String key, int filter, DataCacheChangeCallback delegate);

42. Application
Cache2Cache1
Primary for
K2, V2
Primary for
K1, V1
Cache3
Primary for
K3, V3
Scalable Notifications
AppFabric Caching Client
Routing Table
Register Notification for Key
“K3"
Map Keys
to Partition
Poll
Required Nodes
Nodes Return List of Changes
LSN Order
Partition: P2
Last LSN: 19
Call Delegate
Store Last LSN
Change Log
Partition P1
1 Add K2
2 Del K32
Change Log
(Partition P2)
18 Del K32
19 Del K43
Change Log
33 Add K1
34 Del K22

43. Agenda
The Problem
A Possible Solution
New Possibilities
What to do next

44. Data Center
Pre-Fetch
Hospital

45. Data Center
Pre-Fetch
HospitalRemote Clinic
WAN
Slow!!

46. Data Center
Pre-Fetch
HospitalRemote Clinic
CacheService
WAN

47. Agenda
The Problem
A Possible Solution
New Possibilities
What to do next

48. Web Platform Installer

49. Select Enterprise

50. Install AppFabric

51. Install AppFabric

52. AppFabric on MSDN
http://msdn.microsoft.com/AppFabric

53. Thanks to our contributors