© 2010 IBM Corporation IBM Power Systems Technical University October 18–22, 2010 — Las Vegas, NV Session Title: Designing a PowerHA SystemMirror for AIX High Availability Solution Speaker Name: Michael Herrera Session ID: HA17(AIX)
© 2010 IBM Corporation
IBM Power Systems Technical University
October 18–22, 2010 — Las Vegas, NV
Session Title:Designing a PowerHA SystemMirror for AIX High Availability Solution
Speaker Name: Michael Herrera
Session ID: HA17(AIX)
Michael Herrera ([email protected])
Advanced Technical Skills (ATS)
Certified IT Specialist
Workload-Optimizing Systems
+
Best Practices for Designing a PowerHA
SystemMirror for AIX High Availability Solution
3
Agenda
• Common Misconceptions & Mistakes
• Infrastructure Considerations
• Differences in 7.1
• Virtualization & PowerHA SystemMirror
• Licensing Scenarios
• Cluster Management & Testing
• Summary
4
HACMP is now PowerHA SystemMirror for AIX!
• Current Release: 7.1.0.X– Available on: AIX 6.1 TL06 & 7.1
• Packaging Changes:– Standard Edition - Local Availability– Enterprise Edition - Local & Disaster Recovery
• Licensing Changes:– Small, Medium, Large Server Class
N/AOct 20, 2009PowerHA SystemMirror 6.1.0
N/ANov 14, 2008PowerHA 5.5.0
Sept, 2011Nov 6, 2007HACMP 5.4.1
N/ASept 10, 2010PowerHA SystemMirror 7.1.0
End of Support DateRelease DateVersion
Product Lifecycle:
HA & DR solutions from IBM for your mission-critical AIX applications
* These dates are subject to change per Announcement Flash
5
PowerHA SystemMirror Minimum Requirements
7.1.0.1 – Sep
6.1.0.2 – May 21
5.5.0.6 – June 7
5.4.1.8 – May 13
PowerHA SystemMirror 7.1• AIX 7.1 • AIX 6.1 TL 6 - SP1
PowerHA SystemMirror 6.1• AIX 7.1• AIX 6.1 TL 2 with RSCT 2.5.4.0• AIX 5.3 TL 9, with RSCT 2.4.12.0
PowerHA Version 5.5• AIX 7.1• AIX 6.1 TL2, SP1 and APAR IZ31208, RSCT 2.5.2.0(Async GLVM - APARs IZ31205 and IZ31207)• AIX 5L 5.3 TL 9 with RSCT version 2.4.10.0
HACMP 5.4.1• AIX 6.1 with RSCT version 2.5.0.0 or higher• AIX 5.3 TL4 with RSCT version 2.4.5 (IY84920) or higher• AIX 5.2 TL8 with RSCT version 2.3.9 (IY84921) or higher
6
Common Misconceptions
• PowerHA SystemMirror is an out of the box solution– Scripting & Testing of application Start / Stop scripts– Application monitors will also require scripting & testing
• PowerHA SystemMirror is installed we are completely protected– Consider all single points of failure – ie. SAN, LAN, I/O drawers, etc..
• Heartbeats go over a dedicated link– All interfaces defined to the cluster will pass heartbeats (IP & Non-IP)– CAA definitely changes this behavior
• With clustering I need two of everything – hence idle resources
Fact:
Clustering will highlight what you are & are NOT doing right in your environment
7
Common Mistakes Beyond Base Cluster Functionality
• Not knowing Expected fallover behaviors
• Lack of application monitoring
• Not knowing what to monitor or check– CLI
– Logs
• Not propagating changes appropriately
• No change history
• Down / Missing Serial Networks• EtherChannel Links Down• ERRORs in Verification• Inconsistent AIX levels• Down level cluster filesets• Fallback Policy not set to desired behavior• Missing Filesets• Missing Custom Disk Methods• SAN not built in a robust fashion• Bootlist Issues• Dump Devices
– Insufficient size– Mirrored– Lack Secondary
• I/O Pacing Enabled (old values)• HBA Levels at GA code• Fiber Channel Tunable settings not enabled• Interim Fixes not loaded on all cluster nodes
Lack of Education / Experience
Solutions:
• IBM Training / Redbooks / Proof of Concepts / ATS Health-check Reviews
Poor Change Controls
8
Identify & Eliminate Points of Failure
• LAN Infrastructure
– Redundant Switches
• SAN Infrastructure
– Redundant Fabric
• Application Availability
– Application Monitoring
– Availability Reports
9
Infrastructure Considerations
Important:Identify & Eliminate Single Points of Failure!
Site A Site B
Node A Node B
DWDM DWDM
SITEAMETROVG
50GB
50GB
50GB
50GB
50GB
50GB
50GB
50GB
LAN
SANSAN
LAN
All links through one pipe
10
Infrastructure Considerations
Important:Identify Single Points of Failure & design the solution around them
Site A Site B
Node A
WAN
Node B
XD_rs232
XD_IP
net_ether_0
DWDM DWDM
SITEAMETROVG
1GBECM VG: diskhb_vg1
hdisk2 000fe4111f25a1d1
ECM VG: diskhb_vg1
hdisk3 000fe4111f25a1d1
1GBECM VG: diskhb_vg2
hdisk3 000fe4112f998235
ECM VG: diskhb_vg2
hdisk4 000fe4112f998235
LAN
SAN
LAN
SAN
50GB
50GB
50GB
50GB
50GB
50GB
50GB
50GB
11
Infrastructure Considerations
• Power Redundancy
• I/O Drawers
• SCSI Backplane
• SAN HBAs
• Virtualized Environments
• Application Fallover Protection
I/O drawer
I/O drawer
I/O drawer
I/O drawer
I/O drawer
1
2
3
4
5
6
7
8
9
10
Ie 1. Two nodes sharing I/O drawer
Real Customer Scenarios:
Ie 2. Application Failure with no monitoring – box remains up : no cluster fallover
Moral of the Story:
* High Availability goes beyond just installing the cluster software
12
PowerHA SystemMirror 7.1: Topology managementHeartbeating differences between earlier cluster releases
LPAR 1
LPAR 3
LPAR 2LPAR 4
diskhb_net4 diskhb_net3
diskhb_net1diskhb_net2
RSCT Subnet
Heartbeat Rings
PowerHA SM 6.1 & Earlier
RSCT Based Heartbeating• Leader, Successor, Mayor, etc..• Strict Subnet Rules• No Heartbeating over HBAs
Multiple Disk Heartbeat Networks• Point to Point only• Each Network requires LUN with ECM VG
LPAR 1
LPAR 3
LPAR 2LPAR 4
Multicasting
PowerHA SM 7.1 with CAA
Kernel Based Cluster Message Handling• Multi cast based protocol• Use Network & SAN as needed• Discover and use as many adapters as possible
• All monitors are implemented at low levels of the AIX Kernel & are largely insensitive to system load
Single Repository Disk• Used to heartbeat & store information
13
Transition of PowerHA Topology IP Networks
en1
(persistent IP) 9.19.51.11
( base address) 192.168.100.2en0
9.19.51.20 (service IP)
9.19.51.10 (persistent IP)
192.168.100.1 (base address) en0
en1 ( base address) 192.168.101.2
Traditional heartbeating rules no longer apply. However, route stripping is still a potential issue. When two interfaces have routable IPs on the same subnet AIX will send half the traffic out of either interface
9.19.51.21 (service IP)
192.168.101.1 (base address)
VLAN
Network: Net_ether0
Methods to circumvent this:
• Link Aggregation / EtherChannel• Virtualized Interfaces with dual VIO servers
( base address) 9.19.51.11en2
9.19.51.21 (service IP)
9.19.51.20 (service IP)
9.19.51.10 (base address) en2
VLAN
ent0 ent1 ent0 ent1
HB Rings
In 6.1 & below
14
PowerHA SM 7.1: Additional Heartbeating Differences
Serial Networks removed:• No more rs232 support• No more traditional disk heartbeating over ECM VG• No more slow takeover w/disk heartbeat device as last device on selective takeover
Critical Volume Groups• Replace Multi-node Disk Heartbeating
– Oracle RAC three disk volume group - Voting Files– Unlike MNDHB, no more general use– Migration is a manual operation, and customer responsibility– Any Concurrent Access Volume Group can be marked as “Critical”
( base address) 9.19.51.11en2
9.19.51.21 (service IP)
9.19.51.20 (service IP)
9.19.51.10 (base address) en2
VLAN
ent0 ent1 ent0 ent1
Heartbeating:• Self Tuning Failure Detection Rate (FDR)• All interfaces are used even if not in cluster networks
en3 en3
15
CAA – Cluster Aware AIX
What is it:
• A set of services/tools embedded in AIX to help manage a cluster of AIX nodes and/or help run cluster software on AIX
• IBM cluster products (including RSCT, PowerHA, and the VIOS) will use and/or call CAA services/tools
• CAA services can assist in the management and monitoring of an arbitrary set of nodes and/or running a third-party cluster
• CAA does not form a cluster by itself. It is a tool set.
• There is no notion of quorum (If 20 nodes of a 21 node cluster are down, CAA still runs on the remaining node)
• CAA does not eject nodes from a cluster. CAA provides tools to fence a node but never fences a node and will continue to run on a fenced node
Major Benefits:
• Enhanced Health Management (Integrated Health Monitoring)
• Cluster Wide Device Naming
Enabling tighter Integration with PowerHA SystemMirror
16
Cluster Aware AIX Exploiters
Legacy AIX
PowerHA
System MirrorTSA HMC
IBM
StorageHPC
DB2IBM
Director
Monitoring
API
Cluster
Monitoring
Group Services
Cluster Admin
UI
Cluster CFG
Repository
Resource Mgr Services
Bundled Resource Managers
Cluster
Messaging
Messaging
API
Legacy RSCT
RSCT Consumers
VIOS
Monitoring
API
Cluster
Monitoring
Group Services
Cluster Admin
UI
Cluster CFG
Repository
Resource Mgr Services
Bundled Resource Managers
Cluster
Messaging
Messaging
API
RSCT With Cluster Aware AIX
Cluster Aware AIX
Cluster
Repository
Cluster
Messaging
Cluster
Monitoring
Cluster
Events
CAA APIs and UIs
Redesigned Layers Integrated to CAA Capabilities
• RSCT and Cluster Aware AIX together provide the foundation of strategic Power Systems SW
• RSCT-CAA integration enables compatibility with a diverse set of dependent IBM products
• RSCT integration with CAA extends simplified cluster management along with optimized and robust cluster monitoring, failure detection, and recovery to RSCT exploiters on Power / AIX
17
Cluster Aware AIX: Central Repository Disk
Aids in:
• Global Device Naming• Inter node synchronization• Centrally managed configuration• Heartbeating device
Host 1 Host 2 Host 3
HA ODM HA ODM HA ODM
Cluster Synchronization
PowerHA SystemMirror 6.1 & Prior:
Host 1 Host 2 Host 3
Central
Repository
PowerHA SystemMirror will continue to
run if Central Repository Disk goes away
However, no changes may take place
within the cluster.
PowerHA SystemMirror 7.1 & CAA:
Contrast from previous releases
Direction:
• In the first release, support is confined to shared storage
• Will eventually evolve into a general AIX device rename interface
• Future direction is to enable cluster-wide storage policy settings
• PowerHA ODM will eventually also entirely move to the repository disk
18
Multi Channel Health Management – Out of the Box
Network
SAN
Repository Disk
Heartbeats HeartbeatsReliable
Messaging
Reliable
Messaging
First Line of Defense
Second Line of Defense
Third Line of Defense
Highlights:
• RSCT Topology services no longer used for cluster Heartbeating• All customers now have multiple communication paths by default
Hardened Environments with new communication protocol
Faster detection & more efficient communication
LPAR 1 LPAR 2
19
Basic Cluster vs. Advanced Cluster Features
Basic Cluster
– Network Topology – Resource Group/s
• IPs• VGs• Application Server
– Application Monitoring– Pager Events
Advanced Cluster
– Multiple Networks• Crossover Connections• Virtualized Resources
– Multiple Resource Groups• Mutual Takeover• Custom Resource Groups
– Adaptive Fallover
– NFS Cross-Mounts– File Collections– Dependencies
• Parent / Child• Location• Start After• Stop After
IP
VGs
App Server
IP Network
SAN Network
Resource Group
IP
VGs
App1
IP Network
SAN Network
Resource Group
IP Network
IP
VGs
App2
Resource Group
– Smart Assistants– Multiple Sites– Cross Site LVM Configs– Storage Replication– IP Replication– Application Monitoring– Pager Events– DLPAR Integration
• Grow LPAR on Fallover
– Director Management– WebSMIT Management– Dynamic Node Priority
IP
VGs
Dev App
Resource Group
Site A Site B
Disk Replication
20
One to one One to any
Any to anyAny to one
PowerHA SystemMirror: Fallover PossibilitiesCluster Scalable to 32 nodes
21
Methods to Circumvent Unused Resources
Frame 1
hdisk2
hdisk4
rootvg {
oracle_vg1 {
hdisk1
hdisk4
}rootvg
}oracle_vg1
SAN SAN
VIO
HBA
VIO
Node B
Frame 2
Node A
VIOHBA
VIO
NIC
NIC
NIC
NIC
HBA
HBA
Storage Subsystem
Resource Group A
Node A, Node BShared IPVG/s & filesystemsApp 1
Resource Group B
Node A, Node BShared IPVG/s & filesystemsApp 2
Resource Group C
Node B, Node AShared IPVG/s & filesystemsApp 3
Resource Group D
Node B, Node AShared IPVG/s & filesystemsApp 4
RGDependency
RGDependency
Mutual
Takeover
Virtualization
22
Power Virtualization & PowerHA SystemMirror
• LPAR / DLPAR
• Micropartitioning & Shared Processor Pools
• Virtual I/O Server
– Virtual Ethernet
– Virtual SCSI
– Virtual Fiber
• Live Partition Mobility
• Active Memory Sharing
• WPAR (AIX 6.1)
LAN
SAN
LPAR X
AIXRootvg
Data
vfc1vfc0en0
LPAR Y
VIO1 A VIO2 A
Power HA_node 1
AIXRootvg
Data
vfc1vfc0en0
LPAR Z
VIO1 B VIO2 B
Power HA_node 2
Rootvgvolumes
Data
Power HA Cluster
External
Storage Enclosure
23
PowerHA SystemMirror Virtualization Considerations
• Ethernet Virtualization– Topology should look the same as environment using link aggregation– Version 7.1 no longer uses netmon.cf file– As a best practice dual VIO Servers are recommended
• SEA Fallover Backend
• Storage Virtualization– Both methods of virtualizing storage are supported
• VSCSI vs. Virtual fiber (NPIV)
– In DR implementations leveraging disk replication consider the implications of using either option
• Benefits of virtualization:– Maximize utilization of resources– Less PCI slots & physical adapters– Foundation for advanced functions like Live Partition Mobility– Migrations to newer Power Hardware are simplified
* Live Partition Mobility & PowerHA SM compliment each otherMaintenance vs. High Availability(non-reactive . reactive)
Chapter 2.4 PowerVM Virtualization Considerations
24
Frame 1
Virtual I/O Server (VIOS2)
Virtual Ethernet & PowerHA SystemMirror
Hypervisor
PowerHA LPAR 1
en0
ent0(virt)
Virtual I/O Server (VIOS1)
ent4(SEA)
ent2(virt)
PowerHA LPAR 2
ent0(virt)
en0
ent0(phy)
Ethernet Switch
ent0(phy)
ent4(SEA)
ent2(virt)
Ethernet Switch
ent6(virt)
ent5(virt)
Control
Channel
Control
Channel
ent5(virt)
en6
ent6(virt)
en6
PVID 99
PVID 10
This is a diagram of the configuration required for SEA fallover across VIO Servers. Note that Ethernet traffic will not be load balanced across the VIO Servers. The lower trunk priority on the “ent2” virtual adapter would designate the primary VIO Server to use.
No Link Aggregation / Same Frame
25
Virtual I/O Server (VIOS2)
Virtual Ethernet & PowerHA SystemMirror
Hypervisor
PowerHA LPAR 1
en0
ent0(virt)
Virtual I/O Server (VIOS1)
ent0(phy)
ent4(SEA)
ent2(virt)
ent1(phy)
Ethernet Switch
ent3(LA)
ent0(phy)
ent4(SEA)
ent2(virt)
ent1(phy)
ent3(LA)
Virtual I/O Server (VIOS2)
Hypervisor
PowerHA LPAR 2
en0
ent0(virt)
Virtual I/O Server (VIOS1)
ent0(phy)
ent4(SEA)
ent2(virt)
ent1(phy)
ent3(LA)
ent0(phy)
ent4(SEA)
ent2(virt)
ent1(phy)
ent3(LA)
Ethernet Switch
Frame1
Frame2
ent5(virt)
ent5(virt)
ent5(virt)
ent5(virt)
Control
Channel
Control
Channel
Control
ChannelControl
Channel
Independent Frames & Link Aggregation
26
PowerHA Node 1
PowerHA SystemMirror 6.1 & Below
PowerHA Node 2
en0
FRAME 2
net_ether_0
9.19.51.10
( base address)
9.19.51.11
( base address)
FRAME 1
en0
9.19.51.20 (service IP) (service IP) 9.19.51.21
Topsvcs heartbeating
serial_net_0
Virtual I/O Server (VIOS2)
Hypervisor
AIX Client LPAR
en0
ent0(virt)
Virtual I/O Server (VIOS1)
ent0(phy)
ent4(SEA)
ent2(virt)
ent1(phy)
ent3(LA)
ent0(phy)
ent4(SEA)
ent2(virt)
ent1(phy)
ent3(LA)
FRAME X
ent5(virt)
ent5(virt)
Control
ChannelControl
Channel
* Netmon.cf file used for single adapter networks
27
PowerHA SystemMirror 7.1 - Topology
All nodes are monitored:
Cluster Aware AIX tells you what nodes are in the cluster and information on those nodes including state. A special “gossip”protocol is used over the multicast address to determine node information and implement scalable reliable multicast. No traditional heartbeat mechanism is employed. Gossip packets travel over all interfaces including storage.
Differences:
• RSCT Topology services is no longer used for heartbeat monitoring• Subnet Requirements no longer need to be followed• Netmon.cf file is no longer required or used• All interfaces are used for monitoring even if they are not in an HA network(this may be tunable in a future release)IGMP Snooping must be enabled on the switches
28
LUNSVSCSI
VSCSI Mapping vs. NPIV (virtual fiber)
FRAME 1
Hypervisor
FRAME 2
Node 1
VIOS 2
VIOS 1
LUNS
VIOS 2
VIOS 1
hdisk3
}npiv_vg
STORAGE
SUBSYSTEM
MPIO
NPIV HBA
NPIVHBA
fcs0
fcs1
Hypervisor
Node 2
MPIO
NPIVHBA
NPIVHBA
fcs0
fcs1
vhost0
vhost0
vhost0
vhost0
hdisk4
hdisk3
}npiv_vghdisk4
hdisk1 }vscsi_vghdisk2
hdisk0 } rootvg
hdisk1 }vscsi_vghdisk2
hdisk0 } rootvgMPIO
vscsi0
vscsi1
NPIV
MPIOvscsi0
vscsi1
hdisk
hdisk
hdisk
hdisk
HBA
HBA
HBA
HBA
29
Live Partition Mobility Support with IBM PowerHA
rootvg
Considerations:
• This is a planned move
• It assumes that all resources are virtualized through VIO (Storage & Ethernet connections)
• PowerHA should only experiencea minor disruption to theheartbeats during a move
• IVE / HEA virtual Ethernet is not supported for LPM
• VSCSI & NPIV virtual fiber mappings are supported
VIOS 1
Frame 1 Frame 2
VIOS 2
PowerHANode 1
VIOS 1
VIOS 2
PowerHANode 2
rootvg
datavg
SANStorage
PowerHANode 1
PowerHANode 2
The two solutions compliment each other by providing the ability to perform non-disruptivemaintenance while retaining the ability to fallover in the event of a system or application outage
How does it all work?
30
PowerHA and LPM Feature Comparison
�Automated failover upon vg access loss
�Server Workload Management**
�Automated failover upon any specified AIX error
(via customized error notification of error report entry)
�Automated failover upon HW failure
�Automated failover upon App failure
�Live OS/App move between physical frames*
�Automated failover upon System Failure (OS or HW)
�Software Maintenance
��Hardware Maintenance
�Energy Management**
Live
Partition
Mobility
PowerHA
SystemMirror
*~ 2 seconds of total interruption time** Require free system resources on target system
31
PowerHA SystemMirror: DLPAR Value Proposition
Pros:
• Automated action on acquisition of resources (bound to the PowerHA application server)
• HMC Verification Checking for connectivity to the HMC
• Ability to Grow LPAR on Failover
• Save $ on PowerHA SM Licensing–Thin Standby node
Cons:
• Requires Connectivity to HMC
• Potentially Slower FailoverSystem Specs:– 32-way (2.3GHz) Squad-H+
– 256GB of memory
Results:– 120GB DLPAR add took 1min 55 sec
– 246GB DLPAR add took 4 min 25 sec
– 30% busy running artificial load the add took
4 minutes 36 seconds
• Lacks ability to grow LPAR on-fly
HMCHMC
LPAR A LPAR B
Backup
ssh communication ssh communication
Minimal CPU CountDLPAR CPU Count
Application ServerMinimal CPU Count
32
Oracle DB 1 CPU
Banner DB 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
PeopleSoft 1 CPU
Financial DB 1 CPU
System A System B
Power7 740 – 16 Way
Print Server 2 CPU TSM 2 CPU
+ 1 CPU
Power7 740 – 16 Way
Capacity 10 CPU Capacity 10 CPU
+ 2 CPU
+ 1 CPU
+ 2 CPU
Acquired
via DLPAR
with App
Acquired
via DLPAR
with App
Cluster 1
Cluster 2
Cluster 3
Cluster 4
8 GB2Tivoli Storage Manager 5.5.2.0
32 GB3Production Financial DB
32 GB3Banner Financial DB
4 GB2AIX Print Server
8 GB2Production PeopleSoft
16 GB2Production Oracle DB
MemoryCPUApplications
DLPAR Licensing ScenarioHow does it all work?
33
Oracle DB 1 CPU
Banner DB 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
PeopleSoft 1 CPU
Financial DB 1 CPU
System A System B
Cluster 1
Cluster 2
Cluster 3
Cluster 4
Environment: PowerHA App Server Definitions
Oracle DB 1 CPU
Banner DB 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
PeopleSoft 1 CPU
Financial DB 1 CPU
System A System B
+ 1 CPU
+ 2 CPU
+ 1 CPU
+ 2 CPU
Acquired
via DLPAR
with App
Acquired
via DLPAR
with App
Cluster 1
Cluster 2
Cluster 3
Cluster 4
The actual application requirements are stored in the PowerHA SystemMirror definitions and enforced during the acquisition or release of application server resources
During acquisition of resourcesin the cluster start up the host will ssh to the pre-defined HMC/s to perform the DLPAR operation automatically
Application ServerMin 1Desired 2
HMC
Application ServerMin 1Desired 3
34
Environment: DLPAR Resource Processing Flow
Oracle DB 1 CPU
Banner DB 1 CPU
Standby 1 CPU
Standby 1 CPU
System A System B
DLPAR
Cluster 1
Cluster 2
HMC
- 1 CPU
- 2 CPU
- 1 CPU
- 2 CPU
+ 1 CPU
+ 2 CPU
+ 1 CPU
+ 2 CPU
DLPAR
Oracle DB 2 CPU
Banner DB 3 CPU
Oracle DB 2 CPU
Banner DB 3 CPU
LPAR ProfileMin 1Desired 1Max 2
LPAR ProfileMin 1Desired 1Max 3
Application ServerMin 1Desired 2
Max 2
Application ServerMin 1Desired 3
Max 3
LPAR ProfileMin 1Desired 1Max 2
LPAR ProfileMin 1Desired 1Max 3
1. Activate LPARs Activate LPARs2. Start PowerHA
3. Release resources
Fallover or RG_move
4. Release resources
Stop cluster without takeover
Application ServerMin 1Desired 2
Max 2
Application ServerMin 1Desired 3
Max 3
Read Requirements
Take Aways:
• CPU allocations follow the application server wherever it is being hosted (this model allows you to lower the HA license count)• DLPAR resources will only get processed during the acquisition or release of cluster resources• PowerHA 6.1+ allows provide micro-partitioning support and the ability to also alter virtual processor counts• DLPAR resources can come from free CPUs in shared processor pool or CoD resources
35
Oracle DB 2 CPU
Banner DB 3 CPU
Standby 2 CPU
Standby 3 CPU
Standby 2 CPU
Standby 3 CPU
PeopleSoft 2 CPU
Financial DB 3 CPU
System A System B
Cluster 1
Cluster 2
Cluster 3
Cluster 4
PowerHA SystemMirror: DLPAR Value Proposition
Oracle DB 1 CPU
Banner DB 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
PeopleSoft 1 CPU
Financial DB 1 CPU
System A System B
+ 1 CPU
+ 2 CPU
+ 1 CPU
+ 2 CPU
Acquired
via DLPAR
with App
Acquired
via DLPAR
with App
Cluster 1
Cluster 2
Cluster 3
Cluster 4
PowerHA license counts:
Cluster 1 : 4 CPUsCluster 2 : 6 CPUsCluster 3 : 4 CPUsCluster 4 : 6 CPUs
Total : 20 licenses
HMC
Environment using dedicated CPU model (No DLPAR)
Environment using DLPAR modelPowerHA license counts:
Cluster 1 : 3 CPUsCluster 2 : 4 CPUsCluster 3 : 3 CPUsCluster 4 : 4 CPUs
Total : 14 licenses
36
PowerHA SystemMirror: DLPAR Modified Model
Oracle DB
Banner DB 1 CPU
Standby 1 CPU
Standby 1 CPU
PeopleSoft
Financial DB 1 CPU
System A System B
+ 4 CPU
+ 4 CPU
Acquired
via DLPAR
with App
Acquired
via DLPAR
with App
Cluster 1
Cluster2
HMC
Environment using modified DLPAR modelPowerHA license counts:
Cluster 1 : 6 CPUsCluster 2 : 6 CPUs
Total : 12 licenses
* Consolidated both Prod LPARs into one LPAR. Control separated by Resource Groups
Oracle DB 1 CPU
Banner DB 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
Standby 1 CPU
PeopleSoft 1 CPU
Financial DB 1 CPU
System A System B
+ 1 CPU
+ 2 CPU
+ 1 CPU
+ 2 CPU
Acquired
via DLPAR
with App
Acquired
via DLPAR
with App
Cluster 1
Cluster 2
Cluster 3
Cluster 4
HMCEnvironment using DLPAR model
* Same as previous slide
PowerHA license counts:
Cluster 1 : 3 CPUsCluster 2 : 4 CPUsCluster 3 : 3 CPUsCluster 4 : 4 CPUs
Total : 14 licenses
37
Data Protection with PowerHA SM 7.1 & CAA
ECM VGs were introduced in version 5.1
• Fast Disk Takeover
• Fast Failure Detection
• Disk heartbeating
Disk Fencing in CAA
• Fencing is automatic and transparent
• Cannot be turned off
• Fence group created by cl_vg_fence_init called from node_up
CAA Storage Framework fencing support
• Ability to specify level of disk access allowed by device driver
– Read/Write
– Read Only
– No Access (I/O is held until timeout)
– Fast Failure
Enhanced Concurrent Mode Volume Groups are now required
38
Data Protection with PowerHA SM 7.1 & CAAECM Volume groups and the newly added protection
ACTIVEread/write
PASSIVEread only
/data/app
/data/db
/dataDatavgECM VG
Node A Node B
Shared
LUNs
No Access
Fail all I/Os
In the event
of a failure
on node B
• LVM Enhanced Concurrent Mode VGs (Passive Mode)• Prevent writes to logical volume or volume group devices• Prevent filesystems from being mounted or any change requiring access to it
• CAA Fencing – prevents writes to the disk itself (ie. dd which runs below LVM level)
Cluster
Services
Cluster
Services
read/writeCAA
read onlyCAA
39
Management Console: WebSMIT vs. IBM Director
IBM Systems Director Plug-in
• New for PowerHA SystemMirror 7.1
• Only for management of
7.1 & above
• Same look and feel as IBM suite of products
• Will leverage existing Director implementation
• Uses clvt & clmgr CLI behind the covers
CLI & SMIT sysmirror panels still the most common management interfaces
WebSMIT
• Available since HACMP 5.2
• Required web server to run on host until HACMP 5.5 (Gateway server)
• Did not fall in line with look and feel of other IBM offerings
40
WebSMIT Gateway Model: One-to-Many (6.1 & Below)
WebSMIT converted from a one-to-one architecture to one-to-many
User_1 User_2 User_3 User_4
Cluster_A Cluster_B Cluster_C
*One* WebSMIT server…
managing multiple
clusters…
Multiple WebSMIT users…
accessing multiple clusters…
through *one* WebSMIT
server Standalone
WebSMIT
Server
41
WebSMIT Screenshot: Associations Tab
42
Three tier architecture provides scalability:
User Interface ���� Management Server ���� Director Agent
Director Server
� Central point of control
� Supported on AIX, Linux, and Windows
� Agent manager
User Interface
� Web-based interface
� Command-line interface
PowerHA
AIX
Director
AgentP D
P D
P D
P D
P D
P D
P DDiscovery of clusters
and resources
Director Agent
� Automatically installed on AIX 7.1 & AIX V6.1 TL06
Secure communication
PowerHA SystemMirror Cluster ManagementNew GUI User Interface for Version 7.1 Clusters
43
PowerHA SystemMirror Director Integration
• Accessing the SystemMirror Plug-ins
44
IBM Systems Director: Monitoring Status of Clusters
• Accessing the SystemMirror Plug-ins
45
PowerHA SystemMirror Configuration Wizards
• Wizards
46
PowerHA SystemMirror Smart Assistant EnhancementsDeploy HA Policy for Popular Middleware
47
PowerHA SystemMirror Detailed Views
• SystemMirror Management View
48
IBM Director: Management Dashboard
49
Do you know about clvt & clmgr ?
• “clmgr” announced in PowerHA SM 7.1
– clvt available since HACMP 5.4.1 for Smart Assists
– Hard linked “clmgr” to “clvt”
– Originally clmgr was intended for the Director team & rapidly evolved into a major, unintended, informal line item.
– allows for deviation from clvt without breaking the Smart Assists
• From this release forward, only “clmgr” is supported for customer use
– clvt is strictly for use by the Smart Assists
• New Command Line Infrastructure
– Ease of Management
• Stop
• Start
• Move Resources
# clmgr on cluster � Start Cluster Services on all nodes
# clmgr sync cluster � Verify & Sync Cluster
# clmgr rg appAgroup node=node2 � Move a resource group
50
Do you know about “clcmd” in CAA ?
# lslpp -w /usr/sbin/clcmd
/usr/sbin/clcmd bos.cluster.rte
# clcmd lssrc -g caa
-------------------------------
NODE mutiny.dfw.ibm.com
-------------------------------
Subsystem Group PID Status
clcomd caa 9502848 active
cld caa 10551448 active
clconfd caa 10092716 active
solid caa 7143642 active
solidhac caa 7340248 active
-------------------------------
NODE munited.dfw.ibm.com
-------------------------------
Subsystem Group PID Status
cld caa 4390916 active
clcomd caa 4587668 active
clconfd caa 6357196 active
solidhac caa 6094862 active
solid caa 6553698 active
Allows commands to be run across all cluster nodes
# clcmd lspv
-------------------------------
NODE mutiny.dfw.ibm.com
-------------------------------
hdisk0 0004a99c161a7e45 rootvg active
caa_private0 0004a99cd90dba78 caavg_private active
hdisk2 0004a99c3b06bf99 None
hdisk3 0004a99c3b076c86 None
hdisk4 0004a99c3b076ce3 None
hdisk5 0004a99c3b076d2d None
-------------------------------
NODE munited.dfw.ibm.com
-------------------------------
hdisk0 0004a99c15ecf25d rootvg active
caa_private0 0004a99cd90dba78 caavg_private active
hdisk2 0004a99c3b06bf99 None
hdisk3 0004a99c3b076c86 None
hdisk4 0004a99c3b076ce3 None
hdisk5 0004a99c3b076d2d None
51
PowerHA SystemMirror: Sample Application Monitor
# cat /usr/local/hascripts/ora_monitor.sh#!/bin/kshps –ef | grep ora_pmon_hatest
52
PowerHA SystemMirror: Pager Events
HACMPpager:methodname = "Herrera_notify"desc = “Lab Systems Pager Event"nodename = "connor kaitlyn"dialnum = "[email protected]"
filename = "/usr/es/sbin/cluster/samples/pager/sample.txt"eventname = "acquire_takeover_addr config_too_long event_errornode_down_complete node_up_complete"
retrycnt = 3timeout = 45
# cat /usr/es/sbin/cluster/samples/pager/sample.txtNode %n: Event %e occurred at %d, object = %o
Sample Email:
From: root 09/01/2009Subject: HACMPNode kaitlyn: Event acquire_takeover_addr occurred at Tue Sep 1 16:29:36 2009, object =
Attention:
Sendmail must be working and accessible via the firewall to receive notifications
� Action Taken: Halted Node Connor
53
PowerHA SystemMirror Tunables
• AIX I/O Pacing (High & Low Watermark)– Typically only enable if recommended after performance evaluation
– Historical values 33 & 24 have been updated to 513 & 256 on AIX 5.3 and 8193 & 4096 on AIX 6.1
http://publib.boulder.ibm.com/infocenter/pseries/v5r3/index.jsp?topic=/com.ibm.aix.prftungd/doc/prftungd/disk_io_pacing.htm
• Syncd Setting– Default value of 60 � recommended change to 10
• Failure Detection Rate (FDR) – only for Version 6.1 & below
– Normal Settings should suffice in most environments (note that it can be tuned further)
– Remember to enable FFD when using disk heartbeating
• Pre & Post Custom EVENTs– Entry points for notifications or actions required before phases in a takeover
54
PowerHA SystemMirror: Testing Best Practices
• Test Application scripts and Application monitors thoroughly– Common problems include edits to scripts within scripts
• Test fallovers in all directions– Will confirm start & stop scripts on both locations
• Test Cluster– Lpars within same frame
– Virtual resources
• Utilize Available Tools – Cluster Test Tool
• Testing upgrades “Alternate disk install” is your friend
Best Practice:
Testing should be the foundation for your documentation in the event that someone not PowerHA savvy is there when a failure occurs.
55
How to be successful with PowerHA SystemMirror
• Strict Change Controls– Available test environment
– Testing of changes
• Leverage CSPOC functions– Create / Remove / Change - VGs, LVs, Filesystems
– User Administration
• Know what to look for– cluster.log / hacmp.out / clstrmgr.debug log files
– /var/hacmp/log/clutils.log � Summary of nightly verification
– /var/hacmp/clverify/clverify.log � detailed verification output
munited /# cltopinfo -m
Interface Name Adapter Total Missed Current MissedAddress Heartbeats Heartbeats
--------------------------------------------------------------------------------------------------------------------en0 192.168.1.103 0 0rhdisk1 255.255.10.0 1 1
Cluster Services Uptime: 30 days 0 hours 31 minutes
56
Summary
• Review your infrastructure for potential single points of failure
– Be aware of the potential pitfalls listed in the common mistakes slide
• Leverage Features like:
– File Collections
– Application monitoring
– Pager Notification Events
• Keep up with feature changes in each release
– New dependencies & fallover behaviors
• Virtualizing P7 or P6 environments is the foundation for Live Partition Mobility
– NPIV capable adapters can help simplify the configuration & management
• WebSMIT & IBM Director are the available GUI front-ends
– The cluster release will determine which one to use
57
http://www-03.ibm.com/systems/power/software/availability/aix/index.html
PowerHA SystemMirror IBM Portal
( … or Google ‘PowerHA SystemMirror’ and click I’m Feeling Lucky)
Learn More About PowerHA SystemMirror
Popular Topics:
* Frequently Asked Questions
* Customer References
* Documentation
* White Papers
58
Questions?
Thank you for your time!
59
Additional Resources
• New - Disaster Recovery Redbook
SG24-7841 - Exploiting PowerHA SystemMirror Enterprise Edition for AIX
http://www.redbooks.ibm.com/abstracts/sg247841.html?Open
• New - RedGuide: High Availability and Disaster Recovery Planning: Next-Generation Solutions for Multi server IBM Power Systems Environments
http://www.redbooks.ibm.com/abstracts/redp4669.html?Open
• Online Documentationhttp://www-03.ibm.com/systems/p/library/hacmp_docs.html
• PowerHA SystemMirror Marketing Pagehttp://www-03.ibm.com/systems/p/ha/
• PowerHA SystemMirror Wiki Pagehttp://www-941.ibm.com/collaboration/wiki/display/WikiPtype/High+Availability
• PowerHA SystemMirror (“HACMP”) Redbookshttp://www.redbooks.ibm.com/cgi-bin/searchsite.cgi?query=hacmp