HP 3PAR StoreServ 7200 2-node HP 3PAR Command Line Interface Administrator& - Page 163

Virtualization Software

Page 163 highlights

database of a particular group of users, enhances the performance stability for mail databases for other users and for other applications on the HP 3PAR StoreServ Storage systems. Microsoft Exchange is a highly interactive software, and its users demand a swift response to mouse clicks in the Outlook client. To prevent Microsoft Exchange from reporting errors, be careful to ensure that volumes receive enough IOPs and throughput so that the Microsoft Exchange server delivers sufficiently low I/O response times. Virtualization Software Virtualization platforms, such as VMware vSphere and Microsoft Hyper-V, make use of container files to store one or more virtual hard disk drives used by virtual machines. Each container file is built on one or more LUNs accessed over a SAN. Adding the LUNs that make up the container file to a VVset permits application of a QoS rule. This rule will control the IOPS and/or bandwidth for all virtual machines (VMs) whose virtual hard drives are carved out of that LUN. Be careful to ensure that there is enough bandwidth and IOPS in the QoS rule, so that the applications running on the VMs can deliver acceptable I/O response times. QoS I/O control operates on all VMs sharing a VMware datastore and, in some cases, this level of control may not be granular enough. Recent versions of VMware offer three native types of I/O resource control. Their characteristics, together with those of QoS, are listed in Table 16 (page 163). Table 16 VMware I/O Resource Control Control Mechanism HP 3PAR QoS VMware Storage DRS1 VMware SIOC2 VMware AQD3 I/O control technique: Limit IOPS and bandwidth Migrate VM to other datastore Control queue depth Control queue depth of datastore SAN LUN of datastore SAN LUN in VMkernel; VM in VMkernel shares enforced Reacts on: None I/O latency and space I/O latency utilization Queue Full or Device Busy at LUN or port level Granularity: All VMs in datastore Single VM VV All VMs in a single datastore All hosts using the SAN LUN for the datastore or a particular port on the 3PAR Managed from: HP 3PAR MC VMware vSphere VMware vSphere VMware vSphere Availability: 1, 2, 3 HP 3PAR OS 3.1.2 MU2 vSphere 5.0 and later vSphere 4.1 and later vSphere 3.5 U4 and with Enterprise Plus with Enterprise Plus later with Standard license license license 1 Distributes Resource Scheduler 2 Storage I/O Control 3 Adaptive Queue Depth HP 3PAR Priority Optimization QoS rules operate on volumes inside VVsets on an HP 3PAR StoreServ system. QoS rules are agnostic for the application type they manage. A QoS rule, on a VVset that contains one or more VVs that make up a datastore, controls the I/O across all VMs using that datastore. This could be suboptimal, however, as some VMs need more I/O resources than others. Combining QoS with SIOC offers I/O control up to the level of an individual VM-an I/O share and an optional IOPS limit defined per VM distribute the available I/O capacity in a fair way across VMs, ensuring that no single VM consumes all the I/O provided to the VVset through QoS settings. Note that SIOC will not respond to queue-full messages from QoS directly. HP 3PAR Priority Optimization can also cooperate with AQD to manage I/O. AQD handles I/O congestion in the I/O path to the datastore LUN. It does so by halving the queue length to the LUN Using HP 3PAR Priority Optimization 163

  • 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
  • 123
  • 124
  • 125
  • 126
  • 127
  • 128
  • 129
  • 130
  • 131
  • 132
  • 133
  • 134
  • 135
  • 136
  • 137
  • 138
  • 139
  • 140
  • 141
  • 142
  • 143
  • 144
  • 145
  • 146
  • 147
  • 148
  • 149
  • 150
  • 151
  • 152
  • 153
  • 154
  • 155
  • 156
  • 157
  • 158
  • 159
  • 160
  • 161
  • 162
  • 163
  • 164
  • 165
  • 166
  • 167
  • 168
  • 169
  • 170
  • 171
  • 172
  • 173
  • 174
  • 175
  • 176
  • 177
  • 178
  • 179
  • 180
  • 181
  • 182
  • 183
  • 184
  • 185
  • 186
  • 187
  • 188
  • 189
  • 190
  • 191
  • 192
  • 193
  • 194
  • 195
  • 196
  • 197
  • 198
  • 199
  • 200
  • 201
  • 202
  • 203
  • 204

database of a particular group of users, enhances the performance stability for mail databases for
other users and for other applications on the HP 3PAR StoreServ Storage systems.
Microsoft Exchange is a highly interactive software, and its users demand a swift response to
mouse clicks in the Outlook client. To prevent Microsoft Exchange from reporting errors, be careful
to ensure that volumes receive enough IOPs and throughput so that the Microsoft Exchange server
delivers sufficiently low I/O response times.
Virtualization Software
Virtualization platforms, such as VMware vSphere and Microsoft Hyper-V, make use of container
files to store one or more virtual hard disk drives used by virtual machines. Each container file is
built on one or more LUNs accessed over a SAN. Adding the LUNs that make up the container
file to a VVset permits application of a QoS rule. This rule will control the IOPS and/or bandwidth
for all virtual machines (VMs) whose virtual hard drives are carved out of that LUN. Be careful to
ensure that there is enough bandwidth and IOPS in the QoS rule, so that the applications running
on the VMs can deliver acceptable I/O response times.
QoS I/O control operates on all VMs sharing a VMware datastore and, in some cases, this level
of control may not be granular enough. Recent versions of VMware offer three native types of I/O
resource control. Their characteristics, together with those of QoS, are listed in
Table 16 (page
163)
.
Table 16 VMware I/O Resource Control
VMware AQD
3
VMware SIOC
2
VMware Storage DRS
1
HP 3PAR QoS
Control Mechanism
Control queue depth
of datastore SAN LUN
in VMkernel
Control queue depth
of datastore SAN LUN
in VMkernel; VM
shares enforced
Migrate VM to other
datastore
Limit IOPS and
bandwidth
I/O control technique:
Queue Full or Device
Busy at LUN or port
level
I/O latency
I/O latency and space
utilization
None
Reacts on:
All hosts using the
SAN LUN for the
All VMs in a single
datastore
Single VM
All VMs in datastore
VV
Granularity:
datastore or a
particular port on the
3PAR
VMware vSphere
VMware vSphere
VMware vSphere
HP 3PAR MC
Managed from:
vSphere 3.5 U4 and
later with Standard
license
vSphere 4.1 and later
with Enterprise Plus
license
vSphere 5.0 and later
with Enterprise Plus
license
HP 3PAR OS 3.1.2
MU2
Availability:
1, 2, 3
1
Distributes Resource Scheduler
2
Storage I/O Control
3
Adaptive Queue Depth
HP 3PAR Priority Optimization QoS rules operate on volumes inside VVsets on an HP 3PAR StoreServ
system. QoS rules are agnostic for the application type they manage. A QoS rule, on a VVset that
contains one or more VVs that make up a datastore, controls the I/O across all VMs using that
datastore. This could be suboptimal, however, as some VMs need more I/O resources than others.
Combining QoS with SIOC offers I/O control up to the level of an individual VM—an I/O share
and an optional IOPS limit defined per VM distribute the available I/O capacity in a fair way
across VMs, ensuring that no single VM consumes all the I/O provided to the VVset through QoS
settings. Note that SIOC will not respond to queue-full messages from QoS directly.
HP 3PAR Priority Optimization can also cooperate with AQD to manage I/O. AQD handles I/O
congestion in the I/O path to the datastore LUN. It does so by halving the queue length to the LUN
Using HP 3PAR Priority Optimization
163