MAA - Best Practices for the Cloud

Maximum Availability Architecture
Best Practices for the Cloud
Markus Michalewicz, Senior Director of Product Management
Sebastian Solbach, Consulting Member of Technical Staff

October 25, 2018

Safe Harbor Statement
The following is intended to outline our general product direction. It is intended for
information purposes only, and may not be incorporated into any contract. It is not a
commitment to deliver any material, code, or functionality, and should not be relied upon
in making purchasing decisions. The development, release, timing, and pricing of any
features or functionality described for Oracle’s products may change and remains at the
sole discretion of Oracle Corporation.
3

Program Agenda
Why MAA in the Cloud?
Migrating into the cloud
Cloud MAA Best Practices
Ensure Continuous Service
Summary
1
2
3
4
5
4

Program Agenda
Summary
1
2
3
4
5
5

Oracle Maximum Availability Architecture
Best Practices Blueprints for High Availability
Oracle Maximum Availability Architecture
(MAA) is Oracle's best practices blueprint

•  based on proven Oracle high
availability technologies,
•  expert recommendations
•  and customer experiences.

The goal of MAA is to achieve optimal
high availability for Oracle customers at
the lowest cost and complexity.

https://oracle.com/goto/maa

Copyright © 2018, Oracle and/or its affiliates. All rights reserved. | 7
MAA Evolution – From On-Premises into the Oracle Cloud
On-Premises
On-Premises Exadata and
Recovery Appliance
Database /
Exadata
Cloud
Autonomous
Database
MAA Blueprints and Best Practices
MAA integrated Engineered Systems
(config practices, Exachk, lowest
brownouts, HA QoS, data protection)
Adding MAA Configuration and Life
Cycle Operations, Shifting admin
ownership to Oracle with MAA SLAs

MAA On Third-Party Clouds
8
On-Premises
On-Premises Exadata and
Recovery Appliance
Database /
Exadata
Cloud
Autonomous
Database
MAA Blueprints and Best Practices

MAA integrated Engineered Systems
(config practices, Exachk, lowest
brownouts, HA QoS, data protection)
On Third-Party Clouds
•  Most on-premises MAA Blueprints and Best
Practices can be applied manually
•  You need to consider the cloud architecture
•  No Exadata or RAC are available
•  No MAA optimizations can be performed

You lose most of the benefits

Program Agenda
Summary
1
2
3
4
5
9

Cloud Migration Service ....
Migration into the Cloud
10
Mixed Platforms
Mixed DB Versions
Non-CDB and PDBs
Non-Encrypted (usually)
Oracle Linux
11.2.0.4 (non-CDB)
12.1.0.2, 12.2.0.1, 18.3.0, 19.1 (CDB)
Encrypted
SI, RAC, Exadata

ExaCS
Autonomous
Database
Cloud Service
Simple
Minimal Impact
Near Zero
Downtime

Instantiation Options and Considerations
•  Data Transfer Service for Instantiation
– Data Transfer Disk
– Data Transfer Appliance
•  Migration Connectivity Options
– FastConnect
– IPSec VPN and Internet Gateway
•  Cloud Migration Tools and Documentation
– Zero-Downtime Migration (ZDM)
– Documentation or MOS notes
•  Python script to create a standalone backup of your database
11

Migration Options
•  Physical Migration ensures tablespaces are essentially identical
– Option 1: Cloud RMAN backup/restore or Recovery Appliance backup/restore
– Option 2: Data Guard (< 5 minutes downtime)
– Option 3: Combination of DB Upgrade and PDB Plugin operations
(< 1 hour downtime potential)
•  Logical Migration provides ability to reorganize and to re-optimize
– Cross platform, database and endianness support
– Option 1: Data Pump
– Option 2: GoldenGate (zero downtime potential)
12

Simple Data Guard Migration Steps (MOS 2386116.1)
Source DB Cloud DB
STBY
STBY
STBY
11.2
2. Enable TDE and Encrypt Standby
(optional)
Real-time Encrypted Redo 11.2 11.2
1. Prepare Cloud and Instantiate Cloud Standby Instantiate Standby 11.2 11.2
3. Data Guard Switchover
(5 mins downtime)
SWITCHOVER to CLOUD 11.2 11.2
SWITCHBACK to On-Prem
4. Fallback with Zero Data Loss
(Fallback in 5 mins)

11.2 11.2
13

Zero Downtime Migration (ZDM)
An automated Data Guard migration solution
14
ZDM is expected to support migration using cloud backup/restore and Data Guard by early CY2019. ZDM OTN website coming soon.

Upgrade/Migrate Using GoldenGate
or Data Integration Platform Cloud Services
Cross-endian possible
When to use Migration Process
Source version 8i and later
Migrate from non-CDB to PDB
Changes to database structure
possible
1.  Create cloud instance and PDB
2.  Configure GoldenGate,
GoldenGate or DIPC cloud service
depending on the situation
3.  Synchronize changes
4.  Switch clients to cloud database
Upgrade to new version possible
Minimal downtime migration
GGCS : https://docs.oracle.com/en/cloud/paas/goldengate-cloud/index.html
DIPC https://docs.oracle.com/en/cloud/paas/data-integration-platform-cloud/index.html

ZDM: GoldenGate Hub for Autonomous Database
16
•  Simple and Secure
•  Full Automation of Migration Life Cycle
•  Minimal Overhead for Source
•  Significant Performance Improvements
•  Supplemental logging, Redo Transport
•  Data Pump for Instantiation (ZDM early CY2019)
•  Powerful GoldenGate in the cloud
•  Estimated CY2019
Autonomous Database Cloud
ZDM
Service

Program Agenda
Summary
1
2
3
4
5
17


Zero Data Loss
DR to the Cloud Use Case

Zero Downtime
RAC

Zero Data Loss
Backup to the
Cloud Use Case
Prod/Departmental
Business Critical
Dev, Test, Prod
Mission Critical

Backup and
Recovery
Bronze +

Zero Downtime
High Availability
MAA Reference Architectures Apply in the Cloud
Availability Service Levels are mapped to the Cloud
18
Silver +

Zero Data Loss
HA and DR
GOLD
BRONZE
SILVER
PLATINUM
While all tiers exist in the Oracle cloud, subsequent slides focus on the Bronze, Silver and Gold Levels.

Zero Downtime
Golden Gate Cloud Svc.
Gold +

Zero Downtime
Maintenance / Migration

Single Instance DB
Restartable
Backup/Restore
Silver +
DB Replication with
Active Data Guard

Gold +
GoldenGate
Edition Based Redefinition
Bronze +
Database HA with
Oracle RAC
Application Continuity

MAA Architecture Building Blocks
19
What is available where?
Cloud
Infrastructure
Backup/Restore Options RAC ADG Replication across
ADs/Regions
OCI (BM)
Backup to OCI Object Storage (manual/automatic)

Automatic backup copies
across Availability Domains (ADs)
✓
Across ADs
Across Regions via
VCN peering
OCI (VM) (with SI or RAC) ✓* ✓
Exa-OCI (X6/X7) ✓ ✓
OCC
Backup to NFS, Local Object Storage, ZDLRA on
premise, Cloud Object Storage with tiering
✓ Across ADs and
regions where
available
ExaCC (X6, X7) ✓ ✓
* Oracle RAC VMs are placed in separate Fault Domains.

•  Regions serve different geographies, provide Disaster Recovery
•  Availability Domains provide a High Availability foundation in a Region
•  Fault Domains group hardware and software in an Availability Domain
–  Each Availability Domain has 3 Fault Domains (FDs)
–  OCI will place each Oracle RAC VM in a separate FD
MAA Optimizations in the Oracle Cloud
The Oracle Cloud Infrastructure (OCI) has been optimized to run Oracle RAC
Region 3
Availbility
Domain 1
Availability
Domain 2
Availability
Domain 3
20

MAA Deployment Automation in the Cloud
•  Simple UI / CLI / REST interfaces being configured for MAA topologies
•  Databases are provisioned with MAA parameter configurations
•  MAA made easy in the cloud
– Oracle Cloud Infrastructure (OCI)
– And Cloud at Customer
21
MAA Database Deployment Made Easy
Primary
AD #1
Standby
AD #2
SILVER (HA)
Primary
Region #1
Standby
Region #2
GOLD (DR)
BRONZE
Single
Instance
DB Backup
Service
RAC
SILVER (HA)

Bronze: Single Instance Database with Backups
22
Low Cost MAA Solution for customers that can tolerate higher RTO and RPO
Bronze
Summary
•  Single instance database with backups & auto-
restart capabilities with Oracle Clusterware
•  Optional replication of backup
•  OCI replicates backups across another data center or Availability Domain
•  Restore from backup to resume
service following unrecoverable outages
Features •  Oracle Restart Capabilities (enable with Oracle Clusterware)
•  Multitenant Database with PDB features (12c+ DBs)
•  Online Maintenance (available manually)
•  Corruption Protection (enabled)
•  Flashback Technologies (enabled)
•  Recovery Manager and Cloud Storage
•  Recovery Appliance (future, available for Cloud at Customer)
Single
Instance
Database
Database
Files
Primary
Datacenter
Cloud
Backup
Remote
Datacenter
Replicated
Backups

Bronze - Single Instance Oracle Database
Unplanned Outages and Planned Maintenance
Events Downtime (RTO)
Data Loss
Potential (RPO)
Recoverable Database instance failure Minutes Zero
Recoverable server failure Minutes to hour Zero
Data corruptions, unrecoverable instance, server, database or site
failures
Hours to days
Since last
backup or Near
Zero with RA
Online file move, reorganization/redefinition, and certain patches Zero Zero
Hardware or operating system maintenance and database patches
that cannot be done online
Minutes to hours Zero
Database upgrades: patch sets and full database releases Minutes to hours Zero
Platform migrations Hours to a day Zero
Application upgrades that modify back-end database objects Hours to days Zero
Planned
Maintenance
Unplanned
Outages
23

Cloud MAA – Bronze Level
•  Key customer actions
–  Use Cloud Backup APIs
•  default sectionsize=64GB, RMAN parallelism 4/node, LOW compression)
–  or evaluate ZDLRA with Cloud@Customer for lower RPO
–  Adjust HugePages after adding instances/Databases
–  Use cluster-managed services
•  srvctl add service -d MyDBunqName -s cloudytimes –pdb MyPDB1
–  Leverage online file move and online reorganization when appropriate
•  Operational practices
–  Cloud Test/Dev/UAT resources and real world testing recommended
–  Update DB and security software as recommended
–  Pick up new cloud infrastructure software to leverage new features and incorporate MAA practices
24
Best practices customization

Silver – Option 1: High Availability with Real-Time Failover
25
RTO of Seconds for Server Failures, RPO near Zero with Recovery Appliance
Silver
Summary
•  Active-Active clustering with Oracle RAC
•  All nodes active at all times
•  Real-time failover
Features •  Zero-Downtime rolling maintenance across
Oracle RAC instances (in different Failure Domains)
•  Hardware and OS maintenance
•  Qualified Oracle Database patches
•  Available in the Oracle Cloud and ExaCC
•  Benefit from the
“Application Checklist for Continuous Service”
Remote
Datacenter
Replicated
Backups
RAC
SILVER (HA)
Production – Datacenter #1 or AD1
DR– Datacenter #2 or AD2

Silver Option 1 – High Availability with Real-time Failover
Events Downtime (RTO) Data Loss Exposure (RPO)
Recoverable or unrecoverable RAC instance failure Seconds Zero
Recoverable or unrecoverable server failure Seconds Zero
Data corruptions, unrecoverable database, Availability Domain or
Region failure
Hours to days
Since last backup, or Near-zero with
Recovery Appliance
Online file move, reorganization/redefinition, and patching Zero Zero
Hardware or O.S. maintenance and database patches that can’t
be done online but qualified for RAC rolling install
Zero Zero
App upgrades that modify back-end database objects Hours to days Zero
Planned
Maintenance

Unplanned
Outages

26

Cloud MAA – Silver Level
•  All recommendations for the Bronze Level and:
•  Key customer actions:
– Follow Application Checklist for Continuous Service for RAC
– To lower downtime for instance and server failures,
evaluate a lower FAST_START_MTTR_TARGET setting
– To test for HA or performance impact,
•  one test system should be identical or similar to production
27

Reduce Seconds of Blackout to Minimal Brownout for Exadata DB Software Updates
MAA Optimizations for Reduced Black- and Brownouts
NO Service Draining
TPS vs Time
23 seconds blackout
< 1 sec. brownout
With Automatic Service Draining
& New Exadata Optimizations
TPS vs Time
Exadata only; Oracle Grid Infrastructure 18c (18.3) and higher required

Draining and Failover Locally – Switchover between sites
29 29
Active Data Guard
– Scheduled switchover
– Data Protection, DR
– Query Offload
Data Guard
– Scheduled switchover
– Data Protection, DR
GoldenGate
–  Scheduled switchover
–  Active-active replication
–  Heterogeneous

Sharding
–  Massive OLTP
–  Scheduled switchover
–  Active-active replication
–  Heterogeneous

Fast Application Notification
– Notify draining, failover, load balancing
Transparent Application Continuity
– Application HA
Global Data Services
– Cross Site Placement
RAC
– Online Rolling Maintenance
– Scalability
– Server HA
RAC One
– Online Rolling Maintenance
– Server HA
Production Site
Drain within RAC
Switchover
ADG
Drain within RAC
Production – Datacenter #1 or AD1 DR– Datacenter #2 or AD2

Silver – Option 2: High Availability + DG Fast Start Failover
30
Silver
Summary
•  Single Instance with Data Guard
•  Fast Start Failover protection across ADs
Features •  Use of Active Data Guard is highly recommended
•  to benefit from real time apply & auto block repair

Primary
AD #1
Standby
AD #2
SILVER (HA)

Data Guard Automation for Oracle Cloud Infrastructure
Bare Metal DB
Systems
Virtual Machine DB
Systems
Virtual Machine
RAC DB Systems
Exadata
DB Systems
Data Guard Creation Automatic Manual Manual Manual
Data Guard Role Transitions Automatic Manual Manual Manual
Data Guard Monitoring Automatic Manual Manual Manual
Data Guard Across Regions Manual Manual Manual Manual
Data Guard Fast Start Failover Manual Manual Manual Manual
Primary
AD #1
Standby
AD #2
SILVER (HA)

Silver Option 2 – High Availability with Data Guard Fast Start Failover
32
Events Downtime Data Loss Potential
Recoverable or unrecoverable RAC instance failure Seconds to Min Zero
Recoverable or unrecoverable server failure Seconds to Min Zero
Data corruptions, unrecoverable database, Availability Domain
or Region failure *
Seconds to Min Zero
Online file move, reorganization/redefinition, and patching Zero Zero
Hardware or O.S. maintenance and database patches that can’t
be done online but qualified for RAC rolling install
Minutes to hour Zero
App upgrades that modify back-end database objects Hours to days Zero

Unplanned
Outages

Planned
Maintenance

Gold: Comprehensive HA / DR
33
RTO of Seconds to Minutes, RPO of Zero or Near-Zero
Gold
Summary
•  Active-Active clustering with Oracle RAC
•  All nodes active in each Availability Domain (AD)
•  Real-time failover
•  Real-time data protection, HA & DR using Active
Data Guard
•  Best corruption protection
•  Zero or near-zero data loss
•  Offload read-only and backups
Features •  Minimal Downtime for Database Upgrades using
DBMS_Rolling or transient logical standby
•  Automatic DB failover with potential zero data loss
Primary
Region / AD #1
Standby
Region / AD #2
GOLD (DR)

Gold – Comprehensive Data Protection, HA and DR
Oracle Data Protection
34
Capability Physical Block Corruption Logical Block Corruption
Dbverify,
Analyze
Physical block checks
Logical checks for intra-block and
inter-object consistency
RMAN, ASM Physical block checks Intra-block logical checks
Active Data
Guard
•  Continuous physical block checking at standby
•  Strong isolation to prevent single point of failure
•  Automatic repair of physical corruptions
•  Automatic database failover
•  Detect lost write corruption, auto
shutdown and failover
•  Intra-block logical checks at
standby
Database In-memory block and redo checksum In-memory intra-block checks
ASM Automatic corruption detection and repair using extent pairs
Exadata HARD checks on write, automatic disk scrub and repair HARD checks on write
Runtime Manual

Gold – Comprehensive HA and Data Protection
35
Events Downtime Data Loss Potential
Recoverable or unrecoverable RAC instance failure Seconds (RAC) Zero
Recoverable or unrecoverable server failure Seconds (RAC) Zero
Data corruptions, database unable to restart, site failure
Seconds (Active Data Guard
FSFO*)
Near-zero if ASYNC
Zero if SYNC or FAR SYNC
Online file move, reorganization/redefinition, and
patching
Zero Zero
Hardware or operating system maintenance and database
patches that cannot be done online but are qualified for
RAC rolling install
Zero (RAC) Zero
Database upgrades: patch sets, full database releases Seconds (Active Data Guard) Zero
Platform migrations Seconds (Active Data Guard) Zero
Application upgrades that modify database objects Hours to days Zero
Planned
Maintenance

Unplanned
Outages

* FSFO: Fast-start Failover

Cloud MAA – Gold Level
•  All recommendations for the Bronze and Silver Levels and:
•  Decision making
–  Data Guard FSFO across ADs versus Data Guard FSFO across Regions (Site Failover)
•  Key customer actions:
–  Follow Application Checklist for Continuous Service for Data Guard Fast-Start Failover
–  Data Guard Fast Start setup and tuning failover times is manual (refer to updated Oracle Cloud MAA paper)
–  Database Rolling Upgrade with Data Guard is also manual (see http://oracle.com/goto/maa)
–  Test complete application + Data Guard role transitions
36

Oracle Cloud Infrastructure (OCI) Deployment Model
•  High Availability, data
protection & fast failover
for database across ADs
•  Data Guard synchronous
mode possible due to < 1ms
between ADs
•  Data Guard Fast Start
Failover is manual
•  VCN Peering across Regions
available today
–  PHX – IAD and LHR – FRA
–  Cloud tooling cannot
provision across Regions.
•  No automatic DR
between regions
37
OCI deployment specifics
< 1ms Latency
Latency varies

OCI DG Test Results
38
RPO=0 with SYNC with minimal impact. Fast failover with potential < 30 seconds
Example is MAA tuned; single instance Data Guard Fast-Start Failover on OCI Bare Metal across ADs

MAA Optimizations beyond Gold – for Autonomous
•  Timings are reduced with
Database Release 18 and higher
and for all Autonomous Database
configurations with Data Guard

39
OCI deployment specifics
Example is MAA tuned; single instance Data Guard Fast-Start Failover on OCI Bare Metal across ADs

Program Agenda
Summary
1
2
3
4
5
40

Oracle Clusterware-based services ensure location transparency
Create a Cluster-Managed Service Whenever Possible
[grid@cloudrac1 ~]$ srvctl status database -d myrac_iad18j
Instance myrac1 is running on node cloudrac1
Instance myrac2 is running on node cloudrac2 
[grid@cloudrac1 ~]$ srvctl config service -d myrac_iad18j 
[oracle@cloudrac1 ~]$ . oraenv
ORACLE_SID = [oracle] ? myrac1
ORACLE_HOME = [/home/oracle] ? /u01/app/oracle/product/18.1/dbhome_1
The Oracle base has been set to /u01/app/oracle
[oracle@cloudrac1 ~]$ export ORACLE_UNQNAME=myrac_iad18j
[oracle@cloudrac1 ~]$ srvctl add service -d myrac_iad18j -s cloudytimes -pdb mypdb1 -
preferred myrac1,myrac2
[oracle@cloudrac1 ~]$ srvctl start service -d myrac_iad18j -s cloudytimes
[oracle@cloudrac1 ~]$ srvctl status service -d myrac_iad18j -s cloudytimes
Service cloudytimes is running on instance(s) myrac1,myrac2

Follow “Defining a Custom SCAN Host Name for Exadata Cloud Service” as needed
Use SCAN within the Cloud Automatically
[grid@cloudrac1 ~]$ srvctl config scan
SCAN name: cloudrac-scan.ad1.ssopub.oraclevcn.com,
Network: 1
Subnet IPv4: 10.168.61.0/255.255.255.0/eth0,
static
Subnet IPv6:SCAN 1 IPv4 VIP: 10.168.61.14
SCAN VIP is enabled.
SCAN VIP is individually enabled on nodes:
SCAN VIP is individually disabled on nodes:
SCAN 2 IPv4 VIP: 10.168.61.15
SCAN 3 IPv4 VIP: 10.168.61.13
[grid@cloudrac1 ~]$ srvctl config SCAN_Listener
SCAN Listener LISTENER_SCAN1 exists. Port: TCP:1521
Registration invited nodes:
Registration invited subnets:
SCAN Listener is enabled.
SCAN Listener is individually enabled on nodes:
SCAN Listener is individually disabled on nodes:
Three SCAN IPs / Listeners
per cluster; just as on-premises

Configure SCAN in your own (cloud) DNS for access from the internet
Resolve SCAN Within and Outside of the Cloud
•  One Solution: https://dyn.com/oracle/
•  Any reachable DNS would work
•  More information:
•  https://blogs.oracle.com/cloud-infrastructure/introducing-dns-on-oracle-cloud-infrastructure

Connecting to the database – 3 Options
Choose Your Means of Connection
1.  Local access on one of the DB hosting cloud instances
2.  Access via another cloud instance in the same VCN or via VPN
3.  Access from the internet

Connect to the Database
•  Internal SCAN IP Addresses or SCAN Name (10.168.61.13/14/15)
•  Add 1521 to the Ingress Security List Rules of the subnet
•  Connect
•  Check firewall settings on the DB hosting cloud instances if you have problems (1521)
45
Option 2: Access via another cloud instance in the same VCN or via VPN

[oracle@myclient ~]$ sqlplus system/mypwd@cloudrac-scan.ad1.ssopub.oraclevcn.com:1521/
cloudytimes.ad1.ssopub.oraclevcn.com
SQL*Plus: Release 12.1.0.2.0 Production on Thu Apr 12 15:07:05 2018 
… 
Last Successful login time: Thu Apr 12 2018 14:58:12 +00:00 
Connected to: Oracle Database 18c EE Extreme Perf Release 18.0.0.0.0 - Production

Connect to the Database
•  Only possible if DB system uses a public IP address
•  2 Options:
1.  SQL*Net via. SSH Tunnel
with Public IP Address and VIP address
2.  SQL*Net with Public IP Adress
•  Regard Ingress Rule on Security List for VCN
•  You can set it to 0.0.0.0/0 but that opens it to all internet...
46
Option 3: Access from the internet (sqlcl Example)
SQLCL: http://www.oracle.com/technetwork/developer-tools/sql-developer/downloads/index.html
ss-mac:~ ssolbach$ sql /nolog
SQL> sshtunnel opc@129.213.54.99 -i CloudRSAkey -L
8888:10.168.61.11:1521
Using port:22
SSH Tunnel connected
SQL> connect system/pwd@localhost:8888/
ss-mac:~ ssolbach$ sql system/pwd@129.213.54.99:1521/
SQLcl: Release 18.1 Production on Thu Apr 12 20:35:52
2018 ... 
Connected to: 
Oracle Database 18c EE Extreme Perf Release 18.0.0.0.0 -
Production

It works for the Autonomous Database and all configurations discussed
Follow Application Checklist for Continuous Service

Program Agenda
Summary
1
2
3
4
5
48

Cloud MAA – Customer Benefits
•  MAA and Cloud Development teams collaborate to
– Incorporate MAA configuration Best Practices
– Provide combined and enhanced recommendations:
•  Refer to HA Best Practices guide for an example of the checklist
– Refer to MAA DB Best Practices papers for the overall checklist
•  Refer to Exadata best practices (Exachk - Oracle Exadata Best Practices 757552.1)

– Provide additional recommendations
•  Refer to 888828.1, 1270094.1 and 2333222.1 software recommendations
•  Perform MAA testing and auditing to ensure practices are implemented!
49

No additional customer action required
Example MAA Database Checklist

Oracle Database High Availability in the Cloud
Follow Oracle MAA Blueprints for Oracle Cloud Infrastructure

MAA - Best Practices for the Cloud

More Related Content

What's hot (20)

Similar to MAA - Best Practices for the Cloud (20)

More from Markus Michalewicz (10)

Recently uploaded (20)

MAA - Best Practices for the Cloud