Presented at MariaDB Roadshow Phoenix, December 2017

1. MariaDB MaxScale
James McLaurin
Solution Architect - MariaDB

2. MariaDB MaxScaleMariaDB Multi-Master Cluster
O P E R A T I N G S Y S T E M / F I L E S Y S T E M / S A N / C L O U D
Application
Connectors
MariaDB Server
NoSQL CRUD API
Original Core MariaDB
MariaDB Engineering
Community Contribution
Replicas
Supporting
Asynchronous,
Semi-Sync &
Synchronous
replication
MariaDB
C JDBC ODBC
Replication Kernel Production Plugins
Parallel Slave
GTIDBinLog API
Multi-Source SQL Parser
Cache/Buffer
Optimiser
Connection
Pool
Temporal
PL/SQL
Audit
AWS KMS
Authentication
Handler Socket Etc.
40+ Plugins
SQL
Lightweight Transactional Interoperability
Performance
& Scalability
Graph &
SearchAnalytics
InnoDB
XtraDBAria
Memory
MyISAM
CONNECTColumnStoreSpider OQGRAPH
MyRocks Mroonga
STORAGE LAYER EXTENSIBILITY
KERNEL EXTENSIBILITY

3. 3
MariaDB MaxScale
Latest GA
Version 2.1
Part of MariaDB
TX and AX
Offering
Next Generation
Database Proxy for
SCALABILITY
SECURITY
HIGH AVAILABILITY
DATA STREAMING

4. MariaDB MaxScale Concept
DATABASE
SERVERS
Primary
Secondary
Binlog Cache
Insulates client applications
from the complexities
of backend database cluster
Simplify replication
from database
to other databases
CLIENT
PROTOCOL SUPPORT
AUTHENTICATION
DATABASE MONITORING
LOAD BALANCING & ROUTING
QUERY TRANSFORMATION & LOGGING
Flexible, easy to
write plug-ins for
Generic Core
MULTI-THREADED
E-POLL BASED
STATELESS
SHARES THE THREAD POOL

5. MaxScale Core and Plugins
Filter
Client
Protocol
Protocol
Filter Filter Router
Server State
Monitor
Parser
Core
updates
monitors
uses
Backend

6. Scalability
Transaction Scaling to support user growth
and simplify applications
Connection Rate Limitation
Query Cache
LOAD BALANCING TO
MARIADB MASTER/SLAVE
MARIADB GALERA CLUSTER
AURORA CLUSTER
Replication Scaling to support web-scale
applications’ user base
BINLOG SERVER FOR HORIZONTAL SCALING OF SLAVES
IN MASTER/SLAVE ARCHITECTURE
Multi-tenant database scaling to transparently
grow tenants and data volume SCHEMA SHARDING
HANDLE LARGE AMOUNT OF READ OPERATIONS WITH
HIGH PERFORMANCE

7. Routing based on
QUERY TYPES
QUERY PATTERNS
DATABASE SERVER STATE
REPLICATION LAGS
Load balancing across database clusters
MARIADB MASTER/SLAVE
MARIADB GALERA
MARIADB MULTI-MASTER
AWS AURORA
Scaling
Scale database environment without
application impact
Minimize maintenance downtime
Connection Based Routing
readwrite
Primary
Node
Secondary
read write
Statement Based Routing (Read-
Write Split)
readwrite
Primary Secondary
read &
write

8. Scaling: Tunability
Route all reads to master subsequent to a write for a configurable amount of time or number
of operations
Route queries to only those slave that are less than configurable
replication lag behind master
Route all queries that match a pattern(regex) to specific servers
Weighted routing

9. Scaling: Multi-tenant Database
Each tenant with its own schema
Multi-tenant database hosting
Without impacting existing user base
Scale the database with user base and data volume
growth
Multi-tenant Database Routing
(Schema Sharding)
shard n...shard 2shard 1

10. Scaling: Binlog Replication
Large scale Master/Slave Replication Clusters
Transparent MariaDB binlog replication relay
Horizontally Scale Slaves without Master Overload
Secondary
Primary Node
Secondary
Binlog Cache
Binlog Cache

11. Query Cache
in memory
LRU cache.
Clients
Database Servers
Query Cache Filter
CACHE QUERY RESULTS IN MAXSCALE
FOR CONFIGURABLE TIMEOUT
FOR CACHED QUERIES RETURN
RESULTS FROM CACHE
Handle large amount of read operations
with high performance

12. Security in MaxScale
Black & White List
Connection Rate Limitation
End to End SSL
Database Firewall Filter for SQL Injection protection
Selective Data Masking HIPPA/PCI Compliance
Maximum Rows Returned Limit DDoS Protection
Transport layer security between applications, proxy
& databases
DDoS Protection
LDAP/GSSAPI Authentication Support
Encrypted Binlog Server Files
SSL between binlog server and Master/Slave

13. QUERY FAILED: 1141
ERROR: Required
WHERE/HAVING clause is missing
rule safe_select deny
no_where_clause
on_queries select
rule safe_cust_select deny
regex '.*from.*customers.*'
user %app-user@% match
all rules safe_cust_select
safe_select
Security
DATABASE FIREWALL FILTER
SELECT * FROM CUSTOMERS;
MaxScale
Database Servers
1
2
3
Database Firewall Filter
Allow/Block queries that
MATCH A SET OF RULES
MATCH RULES FOR SPECIFIED USERS
MATCH ON
• date/time
• a WHERE clause
• query type
• column match
• a wildcard or regular expression or function name
Protect against SQL injection
Prevent unauthorized data access
Prevent data damage

14. Security
SELECT Name, creditcardNum, balance
FROM customerTbl
WHERE id=1001
Name creditcardNum balance
---------------------------------------
John Smith xxxxxxxxxx 1201.07
Database Servers
Client
HIPPA/PCI Compliance:
Selective Data Masking
based on column name
DATABASE NAME,
TABLE NAME CLASSIFIER
MAY BE PROVIDED
• commerceDb.customerTbl.creditcardNum
• customerTbl.creditcardNum
• credicardNum

15. Security
DDoS Protection
MAXIMUM ROWS FILTER
• Return zero rows to client if
number of rows in result set
exceeds configured max limit
• Return zero rows to client if
the size of result set exceeds
configured max size in KB Max Rows Limit = 500
NumRows Returned >
MaxRows Limit
QUERY FAILED: 1141
ERROR: No rows returned
51
QUERY
4 MaxRowsLimit FILTER
Clients
NumRows returned = 100032
Database Servers
QUERY

16. Security
Secondary
Primary
Secondary
SSL
SSL
SSL
SSL
Secured Binlog Server
ENCRYPT BINLOG SERVER FILES
on MaxScale
SSL between binlog server
and Master/Slave
Secured user access
LDAP/GSSAPI for secured
single sign-on across OS
platforms(windows, linux),
applications and databases
Binlog Cache Binlog Cache

17. Operational Use Cases
Dynamically configure server, listener, monitor
Dynamically configure database firewall rules
Query Duplication
Query Cache
DUPLICATE QUERY BETWEEN PRODUCTION AND STAGING FOR
UPGRADE VERIFICATION
DUPLICATE QUERY BETWEEN DATABASE AND EXTERNAL
APPLICATIONS FOR QUERY ANALYTICS
Query logging CLUSTER WIDER QUERY TUNING
Query translation TRANSLATE QUERIES SYNTAX FROM ONE PROTOCOL TO ANOTHER
ASYNCHRONOUSLY UPGRADE APPLICATIONS AND DATABASE
INVOKE EXTERNAL OPERATIONAL TOOLS FOR SERVER DOWN/UP
EVENT MANAGEMENT
MANAGE FAILOVER - PROMOTE SLAVE TO MASTER UPON FAILURE

18. Binlog conversion to Avro
Streaming of Change Data Capture
events (binlog) to big data environments
Data Streaming
Change Data Capture
Clients
EMR
MariaDB
ColumnStore
Big Data Platforms
AVRO or
JSON
events
Streaming
MariaDB
Primary
Binlog to AVRO
conversion

19. MariaDB MaxScale 2.2
MariaDB 10.2 Support
REST based management
PAMD Authentication
More performance enhancements
MaxScale CDC to MariaDB ColumnStore Data Adapter

20. Learn more
Download
https://mariadb.com/downloads/mariadb-tx/maxscale
Documentation
https://mariadb.com/kb/en/mariadb-enterprise/maxscale/
Blogs
https://mariadb.com/blog-tags/maxscale
Product Page
https://mariadb.com/products/technology/maxscale
Webinars
https://mariadb.com/resources/webinars?combine=MaxScale
Reach me
jonathan.day@mariadb.com

21. Thank you

22. Example
Read Write Splitting

23. Read/Write Splitting
MariaDB Replication + R/W Split Routing
Each application server uses
only 1 connection
MaxScale monitors the state of each node
and only applies operations on available
slaves
MaxScale creates 2 connections, one for
R/W on the master node and one for
R/O load balanced on the slave nodes
Max
Scale
R/W
Splitting

24. Install MariaDB MaxScale
How to set up
MariaDB MaxScale
Directions here for reference:
https://mariadb.com/kb/en/mariadb-
enterprise/setting-up-mariadb-maxscale/
Install the package
relevant to your
distribution
Step 1
Create the required users
in your MariaDB or
MySQL Replication cluster
Step 2
Create a MariaDB
MaxScale
configuration file
Step 3

25. MaxScale Configuration – Threads
• MariaDB MaxScale configuration is held
in an ini file, /etc/maxscale.cnf.
• This must be manually created, a template file does
exist within the /usr/share/maxscale directory.
• A global, maxscale, section is included within
every MariaDB MaxScale configuration file.
• Various MariaDB MaxScale wide parameters
are set in the file.
• The most important of these is the number of threads
that MariaDB MaxScale will use to execute the code that
forwards requests and handles responses for clients.
[maxscale]
threads=4

26. MaxScale Configuration – Service
• The first step is to create a service
for our Read/Write Splitter.
• Create a section in your MariaDB MaxScale
configuration file and set the type to service.
[Splitter Service]
type=service

27. MaxScale Configuration readwritesplit Module
• The router we need to use for this configuration
is the readwritesplit module.
• The server names given here are actually the names
of server sections in the configuration file and not the
physical hostnames or addresses of the servers.
[Splitter Service]
type=service
router=readwritesplit
servers=dbserv1,
dbserv2, dbserv3

28. MaxScale Configuration – Passwords
• The final step in the service sections is to add the
username and password that will be used to populate
the user data from the database cluster.
• There are two options for representing the password,
either plain text or encrypted passwords.
• In order to use encrypted passwords a set of keys must
be generated that will be used by the encryption and
decryption process. To generate the keys, use the
maxkeys command and pass the name of the secrets
file in which the keys are stored.
maxkeys /var/lib/
maxscale/.secrets

29. MaxScale Configuration – More Passwords
• Once the keys have been created
the maxpasswd command can be used
to generate the encrypted password.
maxpasswd
plainpassword
96F99AA1315BDC3604B00
6F427DD9484

30. MaxScale Configuration – Splitter Service
• The username and password, either encrypted
or plain text, are stored in the service section using
the user and passwd parameters.
[Splitter Service]
type=service
router=readwritesplit
servers=dbserv1, dbserv2, dbserv3
user=maxscale
passwd=96F99AA1315BDC3604B006F427
DD9484

31. MaxScale Configuration – Listener Service
• Listening ports must be associated with the
service in order to allow network connections.
• This is done by creating a series of listener sections.
• A service may have multiple listeners.
[Splitter Listener]
type=listener
service=Splitter Service

32. MaxScale Configuration – More Listener Service
• A listener must also define the protocol module
it will use for the incoming network protocol,
currently this should be the MySQLClient protocol
for all database listeners.
• The listener may then supply a network port to
listen on and/or a socket within the file system.
[Splitter Listener]
type=listener
service=Splitter Service
protocol=MySQLClient
port=3306
socket=/tmp/ClusterMaster

33. MaxScale Configuration – Servers
• An address parameter may be given if the listener
is required to bind to a particular network address
when using hosts with multiple network addresses.
• The next stage is the configuration is to define
the server information.
[dbserv1]
type=server
address=192.168.2.1
port=3306
protocol=MySQLBackend

34. MaxScale Configuration – Servers
• An address parameter may be given if the listener
is required to bind to a particular network address
when using hosts with multiple network addresses.
• The next stage is the configuration is to define
the server information.
[dbserv2]
type=server
address=192.168.2.2
port=3306
protocol=MySQLBackend

35. MaxScale Configuration – Servers
• An address parameter may be given if the listener
is required to bind to a particular network address
when using hosts with multiple network addresses.
• The next stage is the configuration is to define
the server information.
[dbserv3]
type=server
address=192.168.2.3
port=3306
protocol=MySQLBackend

36. MaxScale Configuration – Replication Monitor
• In order for MariaDB MaxScale to monitor the servers
using the correct monitoring mechanisms a section
should be provided that defines the monitor to use and
the servers to monitor.
• Once again a section is created with a symbolic name
for the monitor, with the type set to monitor.
• Parameters are added for the module to use, the list of
servers to monitor and the username and password to
use when connecting to the servers with the monitor.
[Replication Monitor]
type=monitor
module=mysqlmon
servers=dbserv1, dbserv2, dbserv3
user=maxscale
passwd=96F99AA1315BDC3604B006F427
DD9484

37. MaxScale Configuration – Services and Listener
Section
• As with the password definition in the server either
plain text or encrypted passwords may be used.
• The final stage in the configuration is to add the
option service which is used by the maxadmin command
to connect to MariaDB MaxScale for monitoring and
administration purposes. This creates a service section
and a listener section.
[CLI]
type=service
router=cli
[CLI Listener]
type=listener
service=CLI
protocol=maxscaled
socket=default

38. Starting MariaDB MaxScale
• Upon completion of the configuration process
MariaDB MaxScale is ready to be started for the first
time. This may either be done manually by running the
maxscale command or via the service interface.
% maxscale
or
% service maxscale start

39. Services
------------------+----------------+--------+---------------
Service Name | Router Module | #Users | Total Sessions
------------------+----------------+--------+---------------
Splitter Service | readwritesplit | 1 | 1
CLI | cli | 2 | 2
------------------+----------------+--------+---------------
List Services
Check the error log in /var/log/maxscale
to see if any errors are detected in the
configuration file and to confirm
MariaDB MaxScale has been started.
Also the maxadmin command may be
used to confirm that MariaDB MaxScale
is running and the services, listeners etc.,
have been correctly configured.
% maxadmin list services

40. Servers
List Servers
% maxadmin list servers
----------+--------------+-------+-------------+---------------
Server | Address | Port | Connections | Status
----------+--------------+-------+-------------+---------------
dbserv1 | 192.168.2.1 | 3306 | 0 | Running, Slave
dbserv2 | 192.168.2.2 | 3306 | 0 | Running, Master
dbserv3 | 192.168.2.3 | 3306 | 0 | Running, Slave
----------+--------------+-------+-------------+---------------

41. Listeners
List Listeners
MariaDB MaxScale is now ready
to start accepting client connections
and routing them to the master or
slaves within your cluster.
• Other configuration options are available that
can alter the criteria used for routing, these
include monitoring the replication lag within
the cluster and routing only to slaves that are
within a predetermined delay from the
current master or using weights to obtain
unequal balancing operations.
• These options may be found in the MariaDB
MaxScale Configuration Guide. More detail
on the use of maxadmin can
be found in the document.
% maxadmin list listeners
-----------------+-----------------+-----------+-------+--------
Service Name | Protocol Module | Address | Port | State
-----------------+-----------------+-----------+-------+--------
Splitter Service | MySQLClient | * | 3306 | Running
CLI | maxscaled | localhost | 6603 | Running
-----------------+-----------------+-----------+-------+--------

42. Testing the Read - Write Splitting
A simple select statement routes to
one of the slaves
A grouped transaction could contain
writes, so is considered as one. It
should route to the master
MariaDB> select @@hostname;
+------------+
| @@hostname |
+------------+
| dbserve1 |
+------------+
MariaDB> start transaction;
MariaDB> select @@hostname;
+------------+
| @@hostname |
+------------+
| dbserve2 |
+------------+