HP EVA P6550 HP P6000 EVA Updating Product Software Guide (XCS 11200000) (5697 - Page 21

Evaluating your configuration, XCS 11200000 upgrade considerations, Queuing host I/O

Get HP EVA P6550 PDF manuals and user guides View all HP EVA P6550 manuals
Add to My Manuals
Save this manual to your list of manuals

Page 21 highlights

Evaluating your configuration Consider the following factors when determining if an online upgrade is appropriate for your environment: • Queuing host I/O-After the upgrade, the array performs a resynchronization. During the resynchronization, host I/O is queued to the host, which can create a backlog of requests for processing when the array is able to respond. The resynchronization may take less than a minute, but it may take time to clear the backlog of requests and some host I/O may experience timeouts. The HP recommended driver timeout value is 60 seconds. When an I/O timeout occurs, multipath drivers typically retry the I/O request on the same or an alternate path. These retry algorithms have operating system and HBA dependencies and may extend the time that host I/O will be allowed to complete successfully beyond the 60-second timeout. Additionally, some applications may be more sensitive to I/O delays. • Processing host I/O during the upgrade-The upgrade competes with host requests that have a higher priority, which can affect the upgrade outcome. To ensure that host I/O from business critical applications or operations is minimized during the upgrade, select a time when a host I/O disruption would minimize the impact to business operations. • Communicating with HP P6000 Command View-The first steps of the upgrade have a lower priority than host I/O. The upgrade can extend beyond the timeout value expected by HP P6000 Command View, which may cause loss of management connectivity to the array. An HP P6000 Command View discover operation may be required to re-establish communication with the array. XCS 11200000 upgrade considerations Before upgrading to XCS 11200000, consider the following: • Check that the Device Addition Policy is set to manual. If it is set to automatic, you must change it to manual before you perform the controller software upgrade. See "Checking the disk grouping policy" (page 39) for more information. • If the array contains solid state disk (SSD) drives, the upgrade can be done online. • XCS 10100000 and later return disk failure reservation levels to that of versions prior to XCS 10000000. With XCS 10100000 and later, setting the protection level of: ◦ Single consumes twice the capacity of the largest disk group member ◦ Double consumes four times the capacity of the largest disk group member • The EVA resynchronization time is affected by the total virtual mapped capacity. The capacity of all virtual disks, snapshots, snapclones, mirrorclones, containers, data replication log disks, and the number of DR groups contribute to the total virtual mapped capacity. To calculate the values for the formula, use HP P6000 Command View: To calculate (total virtual disk capacity), determine the capacity values for all virtual disks, snapshots, snapclones, mirrorclones, and containers and add them together: ◦ Virtual disks: For each virtual disk, go to the Vdisk - General Properties window and record the value listed in the Requested capacity box. ◦ Snapshots (demand allocated and fully allocated): For each snapshot, go to the Vdisk - General Properties window of the original virtual disk and record the value listed in the Requested capacity box. ◦ Snapclones: For each snapclone, go to the Vdisk - General Properties window of the original virtual disk and record the value listed in the Requested capacity box. Evaluating your configuration 21

  • 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
Evaluating your configuration
Consider the following factors when determining if an online upgrade is appropriate for your
environment:
•
Queuing host I/O
—After the upgrade, the array performs a resynchronization. During the
resynchronization, host I/O is queued to the host, which can create a backlog of requests for
processing when the array is able to respond. The resynchronization may take less than a
minute, but it may take time to clear the backlog of requests and some host I/O may experience
timeouts. The HP recommended driver timeout value is 60 seconds. When an I/O timeout
occurs, multipath drivers typically retry the I/O request on the same or an alternate path.
These retry algorithms have operating system and HBA dependencies and may extend the
time that host I/O will be allowed to complete successfully beyond the 60–second timeout.
Additionally, some applications may be more sensitive to I/O delays.
•
Processing host I/O during the upgrade
—The upgrade competes with host requests that have
a higher priority, which can affect the upgrade outcome. To ensure that host I/O from business
critical applications or operations is minimized during the upgrade, select a time when a host
I/O disruption would minimize the impact to business operations.
•
Communicating with HP P6000 Command View
—The first steps of the upgrade have a lower
priority than host I/O. The upgrade can extend beyond the timeout value expected by HP
P6000 Command View, which may cause loss of management connectivity to the array. An
HP P6000 Command View discover operation may be required to re-establish communication
with the array.
XCS 11200000 upgrade considerations
Before upgrading to XCS 11200000, consider the following:
•
Check that the Device Addition Policy is set to
manual
. If it is set to
automatic
, you must change
it to
manual
before you perform the controller software upgrade. See
“Checking the disk
grouping policy” (page 39)
for more information.
•
If the array contains solid state disk (SSD) drives, the upgrade can be done online.
•
XCS 10100000 and later return disk failure reservation levels to that of versions prior to XCS
10000000. With XCS 10100000 and later, setting the protection level of:
◦
Single
consumes twice the capacity of the largest disk group member
◦
Double
consumes four times the capacity of the largest disk group member
•
The EVA resynchronization time is affected by the total virtual mapped capacity. The capacity
of all virtual disks, snapshots, snapclones, mirrorclones, containers, data replication log disks,
and the number of DR groups contribute to the total virtual mapped capacity.
To calculate the values for the formula, use HP P6000 Command View:
To calculate
(total virtual disk capacity)
, determine the capacity values for all
virtual disks, snapshots, snapclones, mirrorclones, and containers and add them together:
◦
Virtual disks: For each virtual disk, go to the Vdisk – General Properties window and
record the value listed in the Requested capacity box.
◦
Snapshots (demand allocated and fully allocated): For each snapshot, go to the Vdisk –
General Properties window of the original virtual disk and record the value listed in the
Requested capacity box.
◦
Snapclones: For each snapclone, go to the Vdisk – General Properties window of the
original virtual disk and record the value listed in the Requested capacity box.
Evaluating your configuration
21