HP StoreOnce 4430 HP StoreOnce Backup System Concepts and Configuration Guidel - Page 98

Implementing replication with the HP B6200 Backup system

Get HP StoreOnce 4430 PDF manuals and user guides View all HP StoreOnce 4430 manuals
Add to My Manuals
Save this manual to your list of manuals

Page 98 highlights

13 Implementing replication with the HP B6200 Backup system The main difference with the HP B6200 StoreOnce implementation is that replication is part of the service set. Each service set (associated to Node 1, Node 2, et cetera) can handle a maximum of 48 incoming concurrent replication jobs per node and can itself replicate OUT to up to 16 devices. If failover occurs, the replication load becomes incumbent on the remaining service set. The replication traffic will pause during the failover process and restart from the last checkpoint when failover has completed. This means that replication continues without the need for manual intervention but performance may deteriorate. Possible ways to improve this situation are: • Dedicate a couplet as a replication target only (no backup targets). This will allow more resources to be dedicated to replication in the event of failover. • Stagger the replication load across different nodes in different couplets. Try not to overload a couplet that is responsible for replication. Active/Passive and Active/Active configurations Figure 54 shows an ideal situation where: • Site B nodes are acting as replication targets only. Performance is guaranteed and all we have to do is enable the replication windows and make allowances each day for housekeeping. • The replication load at Site B is balanced across two nodes. In the event of failure of a node at Site B, replication performance will not be adversely affected, especially if the nodes at Site B are less than 50% loaded. NOTE: If there are local backups at Site B as well to VTL7 and NAS3, the arrangement shown in Figure 56 would be the best practice. Figure 52shows local backup devices VT7 and NAS7 at Site B on Couplet 1. We are still dedicating nodes to act as replication targets, but they are now on Couplet 2 only. Because the load on the nodes in Couplet 2 is now increased, should a node fail in Couplet 2 on Site B there may be noticeable performance degradation on replication. This is because a single node has to handle a much larger load than in Figure 22. Careful sizing of the nodes in Couplet 2 on Site B to ensure they are less than 50% loaded will ensure that even in failover mode replication performance can be maintained. 98 Implementing replication with the HP B6200 Backup system

  • 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
  • 94
  • 95
  • 96
  • 97
  • 98
  • 99
  • 100
  • 101
  • 102
  • 103
  • 104
  • 105
  • 106
  • 107
  • 108
  • 109
  • 110
  • 111
  • 112
  • 113
  • 114
  • 115
  • 116
  • 117
  • 118
  • 119
  • 120
  • 121
  • 122
13 Implementing replication with the HP B6200 Backup
system
The main difference with the HP B6200 StoreOnce implementation is that replication is part of the
service set. Each service set (associated to Node 1, Node 2, et cetera) can handle a maximum of
48 incoming concurrent replication jobs per node and can itself replicate OUT to up to 16 devices.
If failover occurs, the replication load becomes incumbent on the remaining service set. The
replication traffic will pause during the failover process and restart from the last checkpoint when
failover has completed. This means that replication continues without the need for manual
intervention but performance may deteriorate. Possible ways to improve this situation are:
•
Dedicate a couplet as a replication target only (no backup targets). This will allow more
resources to be dedicated to replication in the event of failover.
•
Stagger the replication load across different nodes in different couplets. Try not to overload
a couplet that is responsible for replication.
Active/Passive and Active/Active configurations
Figure 54 shows an ideal situation where:
•
Site B nodes are acting as replication targets only. Performance is guaranteed and all we
have to do is enable the replication windows and make allowances each day for housekeeping.
•
The replication load at Site B is balanced across two nodes. In the event of failure of a node
at Site B, replication performance will not be adversely affected, especially if the nodes at
Site B are less than 50% loaded.
NOTE:
If there are local backups at Site B as well to VTL7 and NAS3, the arrangement shown
in Figure 56 would be the best practice.
Figure 52shows local backup devices VT7 and NAS7 at Site B on Couplet 1. We are still dedicating
nodes to act as replication targets, but they are now on Couplet 2 only. Because the load on the
nodes in Couplet 2 is now increased, should a node fail in Couplet 2 on Site B there may be
noticeable performance degradation on replication. This is because a single node has to handle
a much larger load than in Figure 22. Careful sizing of the nodes in Couplet 2 on Site B to ensure
they are less than 50% loaded will ensure that even in failover mode replication performance can
be maintained.
98
Implementing replication with the HP B6200 Backup system