PowerHA 7.2.2 Configuration

PowerHA/config2021. 9. 24. 20:41
반응형

 



 
* Written by Jung Byung Wook

구성환경 
- 2NODE : Active - Standby
- Hostname : powerha_01, powerha_02
- Interface IP : 172.30.30.97, 172.30.30.99
- Service IP : 172.30.30.200
- Heartbeat IP : 192.168.10.10, 192.168.10.11
- Data disk : 100G * 4EA
- Repository disk : 30G * 2EA
  

0. 사전 준비단계

 1) PowerHA Best Practice 권장 구성

  • Server Side : Adapter 이중화(HBA, NIC)
  • Server Side : 이더채널 구성(Standard방식 or LACP방식)
  • Server Side : Multipath 구성
  • Network Side : Network Switch 이중화
  • Storage Side : SAN Switch 이중화
  • Storage Side :  Storage Controller port 분산(한 Controller에 집중되지 않게 분산)

Figure1) PowerHA Best Practice 구성도

1) 노드 구성 계획

  • 몇 개의 노드로 구성할 것인지?
  • Active-Standby 구성 혹은 Active-Active(Mutual Takeover) 구성으로 할 것인지?

2) Inter-Connect(Heartbeat) Interface

  • 1G Interface로도 충분하나, 10G 구성도 무방.
  • Heartbeat Interface의 etherchannel은 Active / Standby로 구성하는 것이 일반적.

 3) LUN 구성방식

  • 파일시스템  방식으로 구성할지 혹은 Raw device 방식으로 구성할 것인지?
  • raw device를 사용하는 경우는 concurrent access를 위해 사용함.

 4) 가상화

  • VIOS 구성환경인지 물리적인 환경인지?
  • 구성방법은 다르진 않으나, Network 및 Storage 연결방식상의 차이가 존재.

 5) CAA Repository Disk

  • 기능1 : 노드간 통신을 통해, 클러스터 노드 상태 체크
  • 기능2 : 클러스터 토폴로지 관리(네트워크, 스토리지)
  • 기능3 : 노드간, Resource group migration 정보교환
  • 필요용량 : 1G ~ 460G (일반적으로 30G)
  • 요구사항1 : 공유볼륨
  • 요구사항2 : LUN2개 - Primary / Backup Repository disk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
                                              Cluster Nodes and Networks
 
Move cursor to desired item and press Enter. 
 
  Standard Cluster Deployment
  Multi Site Cluster Deployment                         
  
  Manage the Cluster
  Manage Nodes
  Manage Sites
  Manage Networks and Network Interfaces
  Manage Repository Disks
  =======================
 
  Discover Network Interfaces and Disks
  Add Discovered Network Interfaces to the Configuration
 
  Verify and Synchronize Cluster Configuration
 
 
                                              Manage Repository Disks
 
Move cursor to desired item and press Enter. 
 
  Add a Repository Disk
  =====================
  Remove a Repository disk
  Show Repository Disks
  
  Verify and Synchronize Cluster Configuration
 
                                             Add a Repository Disk
 
Type or select values in entry fields.       
Press Enter AFTER making all desired changes.
 
                                                        [Entry Fields]
* Repository Disk                                     (00c48b554fc68b59)  ==> hdisk2                                                                                                                +
 
------------------------------------------------------------------------------------
                       COMMAND STATUS
Command: OK            stdout: yes           stderr: no
 
Before command completion, additional instructions may appear below.
 
Successfully added one or more backup repository disks.
To view the complete configuration of repository disks use:
"clmgr query repository" or "clmgr view report repository"
 
 cs

1. PowerHA 구성전 기본 설정 및 설치

  •  권장구성방식인 SSH를 통해, GPFS를 구성하기 위한 방식이며, 이 스텝에서는 각 노드가 Password 없이 접근할 수 있게, 노드 상호간의 공개키(Public key)를 공유한다.

 1) /etc/hosts에 상호 노드 등록

  • powerha_01 node : 172.30.30.97
  • powerha_02 node : 172.30.30.99
  • powerha_svc : 172.30.30.200
  • powerha_HB1: 192.168.10.10
  • powerha_HB2: 192.168.10.11

 

 2) PowerHA 패키지 설치

  • 설치 패키지 : cluster.esclient, cluster.es.cspoc, cluster.es.migcheck, cluster.es.nfs, cluster.es.server, cluster.license, cluster.man.en_US.es
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
[powerha_01:root] / > installp -acY -d /home/powerHA cluster.license
[powerha_01:root] / > installp -acY -d /home/powerHA cluster.es.server
[powerha_01:root] / > installp -acY -d /home/powerHA cluster.es.nfs
[powerha_01:root] / > installp -acY -d /home/powerHA cluster.es.migcheck
[powerha_01:root] / > installp -acY -d /home/powerHA cluster.es.cspoc
[powerha_01:root] / > installp -acY -d /home/powerHA cluster.es.client
[powerha_01:root] / > installp -acY -d /home/powerHA cluster.es.assist
 
[powerha_01:root] /home > lslpp -l |grep -i cluster.es
  cluster.es.client.clcomd   7.2.5.0  COMMITTED  Cluster Communication
  cluster.es.client.lib      7.2.5.0  COMMITTED  PowerHA SystemMirror Client
  cluster.es.client.rte      7.2.5.0  COMMITTED  PowerHA SystemMirror Client
  cluster.es.client.utils    7.2.5.0  COMMITTED  PowerHA SystemMirror Client
  cluster.es.cspoc.cmds      7.2.5.0  COMMITTED  CSPOC Commands
  cluster.es.cspoc.rte       7.2.5.0  COMMITTED  CSPOC Runtime Commands
  cluster.es.migcheck        7.2.5.0  COMMITTED  PowerHA SystemMirror Migration
  cluster.es.nfs.rte         7.2.5.0  COMMITTED  NFS Support
  cluster.es.server.diag     7.2.5.0  COMMITTED  Server Diags
  cluster.es.server.events   7.2.5.0  COMMITTED  Server Events
  cluster.es.server.rte      7.2.5.0  COMMITTED  Base Server Runtime
  cluster.es.server.testtool
  cluster.es.server.utils    7.2.5.0  COMMITTED  Server Utilities
  cluster.es.client.clcomd   7.2.5.0  COMMITTED  Cluster Communication
  cluster.es.client.lib      7.2.5.0  COMMITTED  PowerHA SystemMirror Client
  cluster.es.client.rte      7.2.5.0  COMMITTED  PowerHA SystemMirror Client
  cluster.es.cspoc.rte       7.2.5.0  COMMITTED  CSPOC Runtime Commands
  cluster.es.migcheck        7.2.5.0  COMMITTED  PowerHA SystemMirror Migration
  cluster.es.nfs.rte         7.2.5.0  COMMITTED  NFS Support
  cluster.es.server.diag     7.2.5.0  COMMITTED  Server Diags
  cluster.es.server.events   7.2.5.0  COMMITTED  Server Events
  cluster.es.server.rte      7.2.5.0  COMMITTED  Base Server Runtime
  cluster.es.server.utils    7.2.5.0  COMMITTED  Server Utilities
 
 
  

 

 3) /.rhosts & /etc/cluster/rhosts 에 PowerHA 대상 노드 IP 입력

  • 클러스터 통신을 관장하는 CAA Daemon(clcomd)은 rsh을 이용하여, 정보를 주고받는다.
  • inetd daemon 하위 데몬인 rsh의 경우, 보안취약점의 문제로 비활성화하는 경우가 발생하는데, 이는 비권장 사항.
  • rsh이 비활성화된다고 해서, PowerHA 서비스의 문제가 발생하진 않지만, Topology 변경 및 기타 PowerHA의 변경사항이 발생했을 시, 이가 적용되지 않을 수 있다.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
[powerha_01:root] / > cat /.rhosts
172.30.30.97
172.30.30.99
 
[powerha_01:root] / > cat /etc/cluster/rhosts
172.30.30.97
172.30.30.99
 
[powerha_01:root] / > stopsrc -s clcomd
0513-044 The clcomd Subsystem was requested to stop.
[powerha_01:root] / > startsrc -s clcomd
0513-059 The clcomd Subsystem has been started. Subsystem PID is 5963792.
 
[powerha_01:root] / > lssrc -a |grep -i clcomd
 clcomd           caa              5963792      active
 
[powerha_01:root] / > clrsh powerha_02 date
Sat Sep 25 07:07:36 CDT 2021
 
  

 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
[powerha_02:root] / > cat /.rhosts
172.30.30.97
172.30.30.99
 
[powerha_02:root] / > cat /etc/cluster/rhosts
172.30.30.97
172.30.30.99
 
[powerha_02:root] / > stopsrc -s clcomd
0513-044 The clcomd Subsystem was requested to stop.
[powerha_02:root] / > startsrc -s clcomd
0513-059 The clcomd Subsystem has been started. Subsystem PID is 5963792.
 
[powerha_02:root] / > lssrc -a |grep -i clcomd
 clcomd           caa              5963792      active
 
[powerha_02:root] / > clrsh powerha_01 date
Sat Sep 25 07:07:36 CDT 2021
 
  
  • PowerHA 두 노드에서 /etc/cluster/rhosts 파일에 두 노드의 Boot IP(Real IP) 입력.
  • clcomd 데몬 재기동.
  • clrsh 명령어를 통해, 패스워드 없이 각 노드에 접근이 되는지 확인.

 

 4) 디스크 속성 변경

  • reserve_policy : 공유볼륨으로 사용되기때문에, Reservation 정책을 "No_reserve"로 변경.
  • algorithm : 디스크에 발생하는 I/O를 연결된 모든 Path에 Balancing 시켜주는 정책.
  • queue_depth : 디스크 queue size를 16으로 설정
  • 하기 명령어로 변경.#chdev -l hdisk2 -a reserve_policy=no_reserve -a algorithm=round_robin -a queue_depth=16

 

 5) netmon.cf 설정

  • 기능 : PowerHA 구성노드가 아닌 다른 호스트(ex-gateway)의 IP를 지정하여, 해당 호스트로 Ping이 나가지 을 때 네트워크 인터페이스의 Failure을 감지한다.
  • 목적 : 가상화 환경(VIOS)의 경우, Virtual Ethernet Adapter는 항상 UP상태이므로, 특정 호스트로 Ping을 주기적으로 실행하여 인터페이스의 상태를 확인해야하므로, netmon.cf 설정을 필수적으로 적용이 필요하다. 
  • VIOS + PowerHA를 사용하는 경우, 즉각적인 인터페이스의 상태를 확인하기 위해서, Bridge 된 SEA Adapter의 링크 상태를 확인할 수 있는 poll_uplink 를 설정하는 것이 좋다.

          #chdev -l ent0 -a poll_uplink=yes

  • 위치 : /usr/es/sbin/cluster/netmon.cf
1
2
3
4
5
6
7
"Virtual Ethernet Adapter 사용시"
[powerha_01:root] / > vi /usr/es/sbin/cluster/netmon.cf
!REQD en0 172.30.30.1
 
"Physical 환경 혹은 poll_uplink=yes 적용환경"
[powerha_01:root] / > vi /usr/es/sbin/cluster/netmon.cf
172.30.30.1
 
  

2. PowerHA Configuration

 1) Cluster 구성

  • #smitty hacmp > Cluster Nodes And Networks > Standard Cluster Deployment > Setup a Cluster, Nodes and Networks
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
                                             PowerHA SystemMirror
 
Move cursor to desired item and press Enter.
   
  Cluster Nodes and Networks 
  ===========================
  Cluster Applications and Resources
 
  System Management (C-SPOC)
  Problem Determination Tools
  Custom Cluster Configuration
 
  Search for PowerHA SMIT panels
  Learn more about initial cluster configuration
 
                                             Cluster Nodes and Networks
 
Move cursor to desired item and press Enter.
 
  Standard Cluster Deployment
  =============================
  Multi Site Cluster Deployment
 
  Manage the Cluster
  Manage Nodes
  Manage Sites
  Manage Networks and Network Interfaces
  Manage Repository Disks 
  
  Discover Network Interfaces and Disks
  Add Discovered Network Interfaces to the Configuration
 
  Verify and Synchronize Cluster Configuration
 
                                             Standard Cluster Deployment
 
Move cursor to desired item and press Enter.
 
  Setup a Cluster, Nodes and Networks
  ===================================
  Define Repository Disk and Cluster IP Address
 
  Verify and Synchronize Cluster Configuration
  
  Learn more about repository disk and cluster IP address
 
                                             Setup Cluster, Nodes and Networks (Typical)
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
  
                                                     [Entry Fields]
* Cluster Name                                       [powerha_test]  ==> Cluster name
  New Nodes (via selected communication paths)       [powerha_02]    ==> node2 hostname                                                                                                                          +
  Currently Configured Node(s)                        powerha_01
 
 
---------------------------------------------------------------------------------------
                       COMMAND STATUS
 
Command: OK            stdout: yes           stderr: no
 
Before command completion, additional instructions may appear below.
 
[TOP]
 
Warning: to complete this configuration, a repository disk must be defined.
 
clharvest_vg: Initializing....
Gathering cluster information, which may take a few minutes...
clharvest_vg: Processing...
Storing the following information in file
/usr/es/sbin/cluster/etc/config/clvg_config
 
powerha_02:
 
Hdisk:          hdisk0
PVID:           00c48b55df91a889
VGname:         rootvg
VGmajor:        10
Conc-capable:   No
VGactive:       Yes
Quorum-required:Yes
Hdisk:          hdisk1
PVID:           00c48b554fc68a17
VGname:         None
VGmajor:        0
Conc-capable:   No
VGactive:       No
Quorum-required:No
 
 
  

 

2) CAA Repository Disk 등록

  • Node1, Node2에 공유볼륨으로 할당된 hdisk1을 Primary Repository로 지정
  • Cluster Nodes and Networks > Standard Cluster Deployment > Define Repository Disk and Cluster IP Address
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
                                             Standard Cluster Deployment
 
Move cursor to desired item and press Enter.
 
  Setup a Cluster, Nodes and Networks
  
  Define Repository Disk and Cluster IP Address
  =============================================
  Verify and Synchronize Cluster Configuration
  
  Learn more about repository disk and cluster IP address
 
                                             Define Repository and Cluster IP Address
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
  
                                                        [Entry Fields]
* Cluster Name                                        powerha_test
* Heartbeat Mechanism                                 Unicast                                                                                                                             +
* Repository Disk                                    [(00c48b554fc68a17)]  ==> hdisk1                                                                                                                                +
  Cluster Multicast Address                          []
 
 
---------------------------------------------------------------------------------------
                       COMMAND STATUS
 
Command: OK            stdout: yes           stderr: no
 
Before command completion, additional instructions may appear below.
 
Successfully added a primary repository disk.
To view the complete configuration of repository disks use:
"clmgr query repository" or "clmgr view report repository"
Initial site configuration has been saved. You can now go on
to complete the rest of the configuration, including adding backup
repository disks (recommended), custom3 event notifications, resource
groups and applications, etc.
 
When you have entered all the basic information you can then use
Verification and Synchronization to verify the configuration and
distribute it to all cluster nodes.
 
 
  

 

3) Backup Repository disk 등록

  • Cluster Nodes and Networks > Manage Repository Disks > Add a Repository Disk
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
                                             Cluster Nodes and Networks
 
Move cursor to desired item and press Enter. 
 
  Standard Cluster Deployment
  Multi Site Cluster Deployment
  
  Manage the Cluster
  Manage Nodes
  Manage Sites
  Manage Networks and Network Interfaces
  Manage Repository Disks
  ========================
  
  Discover Network Interfaces and Disks
  Add Discovered Network Interfaces to the Configuration
 
  Verify and Synchronize Cluster Configuration
 
Before command completion, additional instructions may appear below.
 
                                             Manage Repository Disks
 
Move cursor to desired item and press Enter. 
 
  Add a Repository Disk
 =======================
  Remove a Repository disk
  Show Repository Disks
  
  Verify and Synchronize Cluster Configuration
 
                                             Add a Repository Disk
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
  
                                                        [Entry Fields]
* Repository Disk                                   [(00c48b554fc68b59)]     ==> hdisk2
  
 
  

 

5) Heartbeat Network 정의

  • Network Heartbeart 기능 :  Node Failure 감지, Node의 지속적인 상태점검, LPM(Live Partition Mobility)간Source / Target Partition간 정보교환.
  • Node의 상태체크를 위한 Heartbeat traffic이다 보니, Traffic의 양이 많지 않다.
  • 서비스 인터페이스/Heartbeat 인터페이스는 별도로 구성하는 것이 권장사항이나, 혼용해서 사용해도 무방.
  • 이 문서에서는 별도의 인터페이스로 구성하는 것을 담았다.

   A) 자동 Discovery된 Heartbeat Interface 제거. (node1 - en1, node2 - en1)

     - Cluster Nodes And Networks > Manage Networks and Network Interfaces > Network Interfaces > Remove a Network Interface > en1 삭제

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
                                             Cluster Nodes and Networks      
 
Move cursor to desired item and press Enter.
 
  Standard Cluster Deployment
  Multi Site Cluster Deployment     
                            
  Manage the Cluster                  
  Manage Nodes                                  
  Manage Sites                
  Manage Networks and Network Interfaces
  =======================================    
  Manage Repository Disks       
                                                
  Discover Network Interfaces and Disks
  Add Discovered Network Interfaces to the Configuration
 
  Verify and Synchronize Cluster Configuration
 
 
                                              Manage Networks and Network Interfaces
 
Move cursor to desired item and press Enter.
 
  Networks                   
  Network Interfaces                
  ==================                    
        
  Show Topology Information by Network
  Show Topology Information by Network Interface
                              
  Verify and Synchronize Cluster Configuration
 
                                             Network Interfaces 
 
Move cursor to desired item and press Enter.
 
  Add a Network Interface    
  Change/Show a Network Interface   
  Remove a Network Interface
  ==========================
                                      
  Verify and Synchronize Cluster Configuration  
 
 
lqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqk
x                  Select Communication Adapter to Remove                  x
x                                                                          x
x Move cursor to desired item and press Enter. Use arrow keys to scroll.   x
x                                                                          x
x   # Node / Network                                                       x
x   #       Interface/Device  IP Label/Device Path              IP Address x
x                                                                          x
x                                                                          x
x   # powerha_01 / net_ether_01                                            x
x           en0               powerha_01                      172.30.30.97 x
x           en1               powerha_HB1                     192.168.10.10x
x           ===============================================================x
x   # powerha_02 / net_ether_01                                            x
x           en0               powerha_02                      172.30.30.99 x
x           en1               powerha_HB2                     192.168.10.11x
x           ===============================================================x
x F1=Help                 F2=Refresh              F3=Cancel                x
x Esc+8=Image             Esc+0=Exit              Enter=Do                 x
x /=Find                  n=Find Next                                      x
mqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqj
  

 

   B) Heartbeat Network 생성

     - Cluster Nodes And Networks > Manage Networks and Network Interfaces > Network  > Add a Network > HB_Network 생성

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
                                             Cluster Nodes and Networks      
 
Move cursor to desired item and press Enter.
 
  Standard Cluster Deployment
  Multi Site Cluster Deployment     
                            
  Manage the Cluster                  
  Manage Nodes                                  
  Manage Sites                
  Manage Networks and Network Interfaces
  =======================================    
  Manage Repository Disks       
                                                
  Discover Network Interfaces and Disks
  Add Discovered Network Interfaces to the Configuration
 
  Verify and Synchronize Cluster Configuration
 
 
                                              Manage Networks and Network Interfaces
 
Move cursor to desired item and press Enter.
 
  Networks                   
  ========                   
  Network Interfaces                
        
  Show Topology Information by Network
  Show Topology Information by Network Interface
                              
  Verify and Synchronize Cluster Configuration
 
                                              Networks 
 
Move cursor to desired item and press Enter.
 
  Add a Network              
  ==============
  Change/Show a Network             
  Remove a Network
                                      
  Verify and Synchronize Cluster Configuration  
 
lqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqk
x                          Select a Network Type                           x
x                                                                          x
x Move cursor to desired item and press Enter.                             x
x                                                                          x
x   ether                                                                  x
x                                                                          x
x F1=Help                 F2=Refresh              F3=Cancel                x
x Esc+8=Image             Esc+0=Exit              Enter=Do                 x
x /=Find                  n=Find Next                                      x
mqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqj
 
                                              Add a Network 
 
Type or select values in entry fields.      
Press Enter AFTER making all desired changes.
                             
                                                        [Entry Fields]
* Network Name                                       [HB_Network]     ==> HB_Network로 지정 
* Network Type                                        ether
* Netmask(IPv4)/Prefix Length(IPv6)                  [255.255.255.0]
* Network attribute                                   private         ==> private로 변경
* Unstable Threshold                                 [3]
* Unstable Period (seconds)                          [60]
 
  

   

  C) Heartbeat Network Interface를 HB_Network에 등록

     - Cluster Nodes And Networks > Manage Networks and Network Interfaces > Network Interfaces  > Add a Network Interface > HB_Network에 "en1" 등록

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
                                             Cluster Nodes and Networks      
 
Move cursor to desired item and press Enter.
 
  Standard Cluster Deployment
  Multi Site Cluster Deployment     
                            
  Manage the Cluster                  
  Manage Nodes                                  
  Manage Sites                
  Manage Networks and Network Interfaces
  =======================================    
  Manage Repository Disks       
                                                
  Discover Network Interfaces and Disks
  Add Discovered Network Interfaces to the Configuration
 
  Verify and Synchronize Cluster Configuration
 
 
                                              Manage Networks and Network Interfaces
 
Move cursor to desired item and press Enter.
 
  Networks                   
  Network Interfaces                
  ==================          
  Show Topology Information by Network
  Show Topology Information by Network Interface
                              
  Verify and Synchronize Cluster Configuration
 
                                              Networks 
 
Move cursor to desired item and press Enter.
 
  Add a Network              
  ==============
  Change/Show a Network             
  Remove a Network
                                      
  Verify and Synchronize Cluster Configuration  
 
 
                                              Network Interfaces 
 
Move cursor to desired item and press Enter.           
                                             
  Add a Network Interface
  ========================                                          
  Change/Show a Network Interface                                     
  Remove a Network Interface  
                                                           
  Verify and Synchronize Cluster Configuration
 
 
lqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqk
x                             Select a Network                             x
x                                                                          x
x Move cursor to desired item and press Enter.                             x
x                                                                          x
x   HB_Network                                                             x
x   ===========                                                            x
x   net_ether_01 (172.30.30.0/24)                                          x
x                                                                          x
x F1=Help                 F2=Refresh              F3=Cancel                x
x Esc+8=Image             Esc+0=Exit              Enter=Do                 x
x /=Find                  n=Find Next                                      x
mqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqj
                                              
                                              Add a Network Interface
 
Type or select values in entry fields.                 
Press Enter AFTER making all desired changes.
                                                                    
                                                        [Entry Fields]
* IP Label/Address                                   [powerha_HB1]     ==> Node Hostname                                                                        
* Network Type                                        ether
* Network Name                                        HB_Network    
* Node Name                                          [powerha_01]      ==> Node name                                                                    
  Network Interface                                  [en1]             ==> Interface name                                                       
 
                                              Add a Network Interface
Type or select values in entry fields.                 
Press Enter AFTER making all desired changes.
                                                                    
                                                        [Entry Fields]
* IP Label/Address                                   [powerha_HB2]     ==> Node Hostname                                                                     
* Network Type                                        ether        
* Network Name                                        HB_Network    
* Node Name                                          [powerha_02]      ==> Node name                                                                      
  Network Interface                                  [en1]             ==> Interface name  
 
 
  

 

6) PowerHA VG 구성

  • Node1에서 VG / LV / FS 생성 
  • Node2에서 Node1에서 생성한 VG의 Major number와 동일하게, Importvg.
  • 생성한 VG는 Automatically 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
"Node1-----------------------------------------------------------------------------"
[powerha_01:root] > lvlstmajor          
38...
[powerha_01:root] > mkvg -V 38 -S -s 128 -a n -y powerha_vg hdisk3 hdisk4 hdisk5 hdisk6
powerha_vg
 
[powerha_01:root] > mklv -t jfs2 -y powerha_lv powerha_vg 10G
 
[powerha_01:root] > crfs -v jfs2 -d powerha_lv -m /powerha -A yes
File system created successfully.
10485236 kilobytes total disk space.
New File System size is 20971520
 
[powerha_01:root] /usr/es/sbin/cluster/etc > varyoffvg powerha_vg
 
"Node2-----------------------------------------------------------------------------"
[powerha_01:root] > lvlstmajor          
38...
[powerha_01:root] > importvg -V 38 -y powerha_vg hdisk3
 
 
  

 

7) Service IP 등록

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
                                             Cluster Applications and Resources
 
Move cursor to desired item and press Enter.
 
  Make Applications Highly Available (Use Smart Assists)
  Resources
  ==============
  Resource Groups
  Cloud Backup Configuration
  
  Verify and Synchronize Cluster Configuration
 
                                             Resources
 
Move cursor to desired item and press Enter.
 
  Configure User Applications (Scripts and Monitors)
  Configure Service IP Labels/Addresses
  =====================================
  Configure Tape Resources
  
  
  Verify and Synchronize Cluster Configuration
 
                                             Configure Service IP Labels/Addresses
 
Move cursor to desired item and press Enter.
 
  Add a Service IP Label/Address
  ==============================
  Change/ Show a Service IP Label/Address
  Remove Service IP Label(s)/Address(es)
  Configure Service IP Label/Address Distribution Preferences
  
  Verify and Synchronize Cluster Configuration
 
  Learn more about Service Labels
 
                                             Add a Service IP Label/Address
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
  
                                                      [Entry Fields]
* IP Label/Address                                    powerha_svc      ==> Service IP 설정
  Netmask(IPv4)/Prefix Length(IPv6)                   []
* Network Name                                        net_ether_01
  

 

8) 기동/종료 스크립트 등록

  • Takeover / Startup시 사용할 기동/종료 스크립트 등록.
  • Cluster Applicatiotns and Resources > Resources > Configure User Applications(Scripts and Monitors) > Application Controller Scripts > Add Application Controller Scripts
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
                                             Cluster Applications and Resources
 
Move cursor to desired item and press Enter.
 
  Make Applications Highly Available (Use Smart Assists)
  Resources
 =================
  Resource Groups
 
  Cloud Backup Configuration
  
  Verify and Synchronize Cluster Configuration
 
                                             Resources
 
Move cursor to desired item and press Enter. 
 
  Configure User Applications (Scripts and Monitors)
  ==================================================
  Configure Service IP Labels/Addresses
  Configure Tape Resources
  
  
  Verify and Synchronize Cluster Configuration
 
                                             Configure User Applications (Scripts and Monitors)
 
Move cursor to desired item and press Enter. 
 
  Application Controller Scripts
  ==================================================
  Application Monitors 
  Resource Optimized High Availability
  
  Show Cluster Applications
 
                                             Application Controller Scripts 
 
Move cursor to desired item and press Enter. 
 
  Add Application Controller Scripts
  ===================================
  Change/Show Application Controller Scripts
  Remove Application Controller Scripts
  
  Learn more about Application Controllers
  
  Verify and Synchronize Cluster Configuration
 
                                             Add Application Controller Scripts
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
  
                                                        [Entry Fields]
* Application Controller Name                        [DB_Script]
* Start Script                                       [/home/dbstart.sh]  ==> Takeover / Startup 시, 기동스크립트
* Stop Script                                        [/home/dbstop.sh]   ==> Takeover / Startup 시, 종료스크립트
  Application Monitor Name(s)
  Application startup mode                           [background]
  Enable CPU usage statistics                         no
  Process to monitor CPU usage                       []
  CPU usage monitor interval                         []
 
  

 

9)  Resource Group 생성

  • Cluster Applications and Resources > Resource Groups > Add a Resource Group
  • Resource Group name 지정
  • Node 우선순위는 지정한 순서대로 지정됨 : 1순위 - powerha_01, 2순위 - powerha_02   
  • Startup Policy : Online On Home Node Only 

     ▷ 클러스터 노드가 down상태에서 시작했을 때, 우선순위가 높은 노드에서만 Resource Group Bring up.

  • Fallover Policy : Failover To Next Priority Mode in The List

     ▷ Fallover(=Takeover)될 때, 다음 우선순위의 노드로 Fallover

  • Fallback Policy : Nevel Fallback

     ▷ Fallover(=TakeoveR)된 이후, 다른 노드가 정상화되더라도, 다시 원복하지 않음.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
                                             Cluster Applications and Resources
 
Move cursor to desired item and press Enter.
 
  Make Applications Highly Available (Use Smart Assists)
 
  Resources
 
  Resource Groups
 ================= 
  Cloud Backup Configuration
  
  Verify and Synchronize Cluster Configuration
                                             Resource Groups
 
Move cursor to desired item and press Enter.
 
  Add a Resource Group
  =====================
  Change/Show Nodes and Policies for a Resource Group   
  Change/Show Resources and Attributes for a Resource Group
  Remove a Resource Group
  Configure Resource Group Run-Time Policies
  Show All Resources by Node or Resource Group        
  
  Verify and Synchronize Cluster Configuration        
  
  Learn more about Resource Groups
 
 
                                             Add a Resource Group
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
  
                                                        [Entry Fields]
* Resource Group Name                                [HA_RG]
* Participating Nodes (Default Node Priority)        [powerha_01 powerha_02]
  
  Startup Policy                                      Online On Home Node Only
  Fallover Policy                                     Fallover To Next Priority Node In The List                                                                                     
  Fallback Policy                                     Never Fallback
 
  

 

10) Resource Group 속성 추가

  • Cluster Applications and Resources > Resource Groups > Change/Show Resources and Attributes for a Resource Group > HA_RG 선택
  • Resource Group에 Servce IP, 기동/종료 스크립트, HA VG 추가.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
                                             Resource Groups
 
Move cursor to desired item and press Enter.
 
  Add a Resource Group
  Change/Show Nodes and Policies for a Resource Group   
  Change/Show Resources and Attributes for a Resource Group
  ==========================================================
  Remove a Resource Group
  Configure Resource Group Run-Time Policies
  Show All Resources by Node or Resource Group        
  
  Verify and Synchronize Cluster Configuration        
  
  Learn more about Resource Groups
 
                                             Change/Show All Resources and Attributes for a Custom Resource Group
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
  
[TOP]                                                   [Entry Fields]
  Resource Group Name                                 HA_RG
  Participating Nodes (Default Node Priority)         powerha_01 powerha_02
  
  Startup Policy                                      Online On Home Node Only
  Fallover Policy                                     Fallover To Next Priority Node In The List
  Fallback Policy                                     Never Fallback
  
  Service IP Labels/Addresses                        [powerha_svc]   ==> Service IP 지정
  Application Controllers                            [DB_Script]     ==> 기동/종료 스크립트 지정
  
  Volume Groups                                      [powerha_vg]    ==> HA VG 
  Use forced varyon of volume groups, if necessary    false
  Automatically Import Volume Groups                  false
 
  Filesystems (empty is ALL for VGs specified)       []
  Filesystems Consistency Check                       fsck
  Filesystems Recovery Method                         sequential
  Filesystems mounted before IP configured            false
                                                        
  Filesystems/Directories to Export (NFSv2/3)        []
  Filesystems/Directories to Export (NFSv4)          []
  Stable Storage Path (NFSv4)                        []
  Filesystems/Directories to NFS Mount               [] 
  Network For NFS Mount                              []
                                                        
  Tape Resources                                     []
 
  

 

11) Cluster 정보 동기화 및 설정 검증

  • Custom Cluster Configuration > Verify and Synchronize Cluster Configuration (Advanced)
  • Cluster, Network, Repository 디스크의 생성 및 선언에 대한 검증.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
                                             PowerHA SystemMirror
 
Move cursor to desired item and press Enter.
 
  Cluster Nodes and Networks
  Cluster Applications and Resources
  
  System Management (C-SPOC)
  Problem Determination Tools
  Custom Cluster Configuration
  ==============================
 
  Search for PowerHA SMIT panels
  Learn more about initial cluster configuration
 
                                             Custom Cluster Configuration
 
Move cursor to desired item and press Enter.
 
  Cluster Nodes and Networks
  Sites
  Resources
  Resource Groups
  Events
  Verify and Synchronize Cluster Configuration (Advanced)
  ========================================================
 
                                             PowerHA SystemMirror Verification and Synchronization
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
  
                                                        [Entry Fields]
* Verify, Synchronize or Both                        [Both]                                                               
* Include custom verification library checks         [Yes]                                                                
* Automatically correct errors found during          [Yes]   ==> 에러 자동수정. YES 
  verification?
 
* Force synchronization if verification fails?       [No]                                                                 
* Verify changes only?                               [No]                                                                 
* Logging                                            [Standard]                                                           
* Detailed checks                                     No                                                                  
* Ignore errors if nodes are unreachable ?            No                                                                  
  

 

10) PowerHA Cluster 기동.

  • #smitty cl_start
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
[powerha_01:root] / > smitty cl_start
                                             Start Cluster Services
 
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
 
                                                        [Entry Fields]
* Start now, on system restart or both                now                                                                                                                              
  Start Cluster Services on these nodes              [powerha_01,powerha_02] 
* Manage Resource Groups                              Automatically
  BROADCAST message at startup?                       false
  Startup Cluster Information Daemon?                 true       ==> Cluster 정보를 출력해주는 데몬 기동
  Ignore verification errors?                         false 
  Automatically correct errors found during           Yes        ==> 기동시, 에러 자동수정 - YES변경
  cluster start?
 
  

 

3. PowerHA 구성확인 및 데몬점검

  • 구성완료 후, PowerHA 데몬시작이 정상적으로 완료되었으면, 정상 구성완료.

 1) PowerHA Cluster Daemon 상태 확인

  • #lssrc -ls clstrmgrES|grep -i state   

          -  Current state: ST_STABLE 

 

 2) Active Node에서 파일시스템 마운트 여부 확인

  • #df -gP
  • #lspv

         - Active Node에서 Resource group 포함한 VG의 hdisk가 "Concurrent"인지 확인.

 

1
2
3
4
5
6
7
8
[powerha_01:root] /usr/es/sbin/cluster > lspv
hdisk0          00c48b55df91a319                    rootvg          active      
hdisk1          00c48b554fc68a17                    caavg_private   active      
hdisk2          00c48b554fc68b59                    None                        
hdisk3          00c48b551d0d870e                    powerha_vg      concurrent  
hdisk4          00c48b551d0d85fe                    powerha_vg      concurrent  
hdisk5          00c48b551d0d84e8                    powerha_vg      concurrent  
hdisk6          00c48b551d0d83d2                    powerha_vg      concurrent 
  

 

 3) Active Node에서 Service IP alias 확인. 

  • #ifconfig -a
1
2
3
4
5
6
7
8
9
10
11
12
[powerha_01:root] /usr/es/sbin/cluster > ifconfig -a
en0: flags=1e084863,10480<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT,CHECKSUM_OFFLOAD(ACTIVE),CHAIN>
        inet 172.30.30.200 netmask 0xffffff00 broadcast 172.30.30.255
        inet 172.30.30.97 netmask 0xffffff00 broadcast 172.30.30.255
         tcp_sendspace 262144 tcp_recvspace 262144 rfc1323 1
en1: flags=1e084863,480<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT,CHECKSUM_OFFLOAD(ACTIVE),CHAIN>
        inet 192.168.10.10 netmask 0xffffff00 broadcast 192.168.10.255
         tcp_sendspace 262144 tcp_recvspace 262144 rfc1323 1
lo0: flags=e08084b,c0<UP,BROADCAST,LOOPBACK,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT,LARGESEND,CHAIN>
        inet 127.0.0.1 netmask 0xff000000 broadcast 127.255.255.255
        inet6 ::1%1/0
         tcp_sendspace 131072 tcp_recvspace 131072 rfc1323 1
 
  

 

 4) Resource Group 상태 확인

  • #/usr/es/sbin/cluster/utilities/clRGinfo
1
2
3
4
5
6
[powerha_01:root] /usr/es/sbin/cluster/utilities > ./clRGinfo
-------------------------------------------------------------------------------------------------------------------------------
Group Name                   Group State      Node                                                             
-------------------------------------------------------------------------------------------------------------------------------
HA_RG                        ONLINE           powerha_01                                                       
                             OFFLINE          powerha_02   
 
  

 

 5) Repository 구성확인

  • #/usr/es/sbin/cluster/utilities/clmgr query repository
1
2
3
[powerha_01:root] /usr/es/sbin/cluster/utilities > ./clmgr query repository
hdisk1 (00c48b554fc68a17)
hdisk2 (00c48b554fc68b59)
 
  
  •  
 
반응형
저작자표시
Posted by 정병욱

티스토리툴바