Guidelines for Using Volume Shadowing Parameters |
 |
This section provides guidelines for using volume
shadowing parameters.
ALLOCLASS
The ALLOCLASS parameter is used to specify an allocation
class that forms part of a device name. The purpose of allocation
classes is to provide unique and unchanging device names. When using
Volume Shadowing for OpenVMS on a single system or on an OpenVMS Cluster
system, a nonzero allocation class value is required for each physical
device in the shadow set. For more information about using allocation
classes, see the HP OpenVMS Cluster Systems manual.
SHADOWING
The SHADOWING parameter enables
or disables volume shadowing on your system, as shown in Table 3-2.
Table 3-2 SHADOWING Parameter Settings
| Setting | Effect |
|---|
0 | Shadowing is not enabled.
This is the default value. |
2 | Enables host-based shadowing. This setting provides shadowing of all disks that are located on
a standalone system or on an OpenVMS Cluster system. Set SHADOWING
to 2 on every node that mounts a shadow set, including satellite nodes. |
SHADOW_HBMM_RTC (Integrity servers and Alpha)
SHADOW_HBMM_RTC is used to
specify, in seconds, how frequently each shadow set on this system
has its modified block count compared with its reset threshold. If
the modified block count exceeds the reset threshold, the bitmap for
that shadow set is zeroed. This comparison is performed for all shadow
sets mounted on the system that have HBMM bitmaps. The reset threshold
is specified by the RESET_THRESHOLD keyword in the /POLICY qualifier
of the SET SHADOW command. When the comparison
is made, the modified block count might exceed the reset threshold
by a small increment or by a much larger amount. The difference depends
on the write activity to the volume and the setting of this parameter.
The default setting of SHADOW_HBMM_RTC
is 150 seconds.
You can view the reset threshold setting and the
modified block count, since the last reset, in the SHOW SHADOWcommand display. For guidelines on setting the reset threshold values
and a sample SHOW SHADOW display, see “Considerations for Setting a Bitmap RESET_THRESHOLD Value”.
For a SHOW SHADOW display that includes a modified
block count greater than the reset threshold value, see the Example 9 for SHOW SHADOW in the HP OpenVMS DCL Dictionary.
SHADOW_MAX_COPY
The SHADOW_MAX_COPY parameter controls how many
parallel copy and merge operations are allowed on a given node. (Copy
and merge operations are described in Chapter 6.) This parameter provides a way to limit the
number of copy and merge operations in progress at any time.
The value of SHADOW_MAX_COPY can range from 0
to 200. The default value is specific to the OpenVMS version. You
can determine the default value by looking at the parameter setting.
When the value of the SHADOW_MAX_COPY parameter is 4, and you mount
five multivolume shadow sets that all need a copy operation, only
four copy operations can proceed. The fifth copy operation must wait
until one of the first four copies completes.
Consider the following when choosing a value for
the SHADOW_MAX_COPY parameter:
Disk controller bandwidth
Interconnect controller
bandwidth
Other work loads on the
system
For example, the default value of 4 may be too
high for a small node. (In particular, satellite nodes must have SHADOW_MAX_COPY
set to a value of 0.) Too low a value for SHADOW_MAX_COPY unnecessarily
restricts the number of operations your system can effectively handle
and extends the amount of time it takes to merge all of the shadow
sets.
SHADOW_MAX_COPY is a dynamic parameter. Changes
to the parameter affect only future copy and merge operations; current
operations (pending or already in progress) are not affected.
SHADOW_MAX_UNIT
The SHADOW_MAX_UNIT specifies the number of shadow sets
that can exist on a node and determines the memory reserved for the
bitmap for each shadow set. (See “Memory Requirements”.) The important thing to note about
this value is that any shadow set that has been created, regardless
of whether it is in use, is included in this total. Because this is
not a dynamic system parameter, you must be very careful when determining
the value to use. If you have to change this parameter, you must reboot
the system.
The default value for OpenVMS Alpha systems is
500.
|
| |
|
 | CAUTION: Any MOUNT command that attempts to create more
shadow sets than the maximum specified for the node fails. |
|
| |
|
Note that this parameter does not affect the naming
of shadow sets. For example, with the default value of 100, a device
name such as DSA999 is still valid.
SHADOW_MBR_TMO
The SHADOW_MBR_TMO
parameter controls the amount of time the system tries to fail over
physical members of a shadow set before removing them from the set.
SHADOW_MBR_TMO is a dynamic parameter that you can change on a running
system.
With the SHADOW_MBR_TMO parameter, you specify
the number of seconds, from 1 to 65,535, during which recovery of
a shadow set member is attempted.
|
| |
|
 | NOTE: The value of SHADOW_MBR_TMO must not exceed the
value of the parameter MVTIMEOUT. |
|
| |
|
If you specify zero, a default delay is used.
The default delay is specific to the version of OpenVMS running on
your system. For shadow sets in an OpenVMS Cluster configuration,
the value of SHADOW_MBR_TMO must be set to the same value on each
node.
Determining the correct value for SHADOW_MBR_TMO
is a trade-off between rapid recovery and high availability. If rapid
recovery is required, set SHADOW_MBR_TMO to a low value. This ensures
that failing shadow set members are removed from the shadow set quickly
and that user access to the shadow set continues. However, removal
of shadow set members reduces data availability and, after the failed
member is repaired, a full copy operation is required when it is mounted
back into the shadow set.
If high availability is paramount, set SHADOW_MBR_TMO
to a high value. This allows the shadowing software additional time
to regain access to failed members. However, user access to the shadow
set is stalled during the recovery process. If recovery is successful,
access to the shadow set continues without the need for a full copy
operation, and data availability is not degraded. Setting SHADOW_MBR_TMO
to a high value may be appropriate when shadow set members are configured
across LANs that require lengthy bridge recovery time.
Shadowing uses a timer to adhere to the number
of seconds specified by the SHADOW_MBR_TMO parameter. For directly
connected SCSI devices that have been powered down or do not answer
to polling, the elapsed time before a device is removed from a shadow
set can take several minutes.
The use of default settings for certain system
parameters may lead to the occasional removal of shadow set members
(systems that are using Volume Shadowing for OpenVMS) that are configured
for multi-path support. Therefore, when configuring multi-path shadow
sets using Volume Shadowing for OpenVMS, follow the recommendations
shown in Table 3-3.
Table 3-3 System Parameter Settings for Multipath Shadow Sets
System Parameter | Recommended Setting |
|---|
MSCP_CMD_TMO | 60 as a minimum. The value of
60 is appropriate for most configurations. Some configurations may
require a higher setting. |
SHADOW_MBR_TMO | At least 3 x MSCP_CMD_TMO |
SHADOW_SYS_TMO | At least 3 x MSCP_CMD_TMO |
MVTIMEOUT | At least 4 x SHADOW_MBR_TMO |
|
| |
|
| NOTE: The recommended setting for MVTIMEOUT, as shown
in Table 3-3, represents a doubling
of an earlier recommendation published for OpenVMS Alpha Version 7.3. |
|
| |
|
To modify SHADOW_MBR_TMO for an existing shadow
set member, see the SET SHADOW/RECOVERY_OPTIONS=DELAY_PER_SERVED_MEMBER=n command, which is described in “Managing Copy and Merge Operations (Integrity servers and Alpha)”.
SHADOW_PSM_DLY
SHADOW_PSM_DLY allows the system manager to adjust the
delay that Shadowing adds automatically when a copy or merge operation
is needed on a shadow set
that is mounted on many systems.
The Shadowing facility attempts
to perform the operation on a system that has a local connection to
all the shadow set members. Shadowing implements the copy or merge
operation by adding a time delay based on the number of shadow set
members that are MSCP-served to the system. No delay is added for
local members. Therefore, a system with all locally accessible shadow
set members usually performs the copy or merge before a system on
which one or more members is served and is therefore delayed.
When a shadow
set is mounted on a system, the value of SHADOW_PSM_DLY is used as
the default shadow set member recovery delay for that shadow set.
To modify SHADOW_PSM_DLY for an existing shadow set, see the SET SHADOW/RECOVERY_OPTIONS=
DELAY_PER_SERVED_MEMBER=n command, which is described in “Managing Copy and Merge Operations (Integrity servers and Alpha)”.
SHADOW_PSM_DLY is a static
parameter; its range is 0 to 65535 seconds. The default value is 30
seconds for each MSCP served shadow set member.
SHADOW_REC_DLY
SHADOW_REC_DLY governs the
system behavior after a system failure or after a shadow set is aborted. The value of the SHADOW_REC_DLY parameter is added to
the value of the RECNXINTERVAL parameter to determine
how long a system waits before it attempts to manage a merge or copy
operation on any shadow sets that it has mounted.
SHADOW_REC_DLY can be used
to predict the systems that can perform recovery operations in an
OpenVMS Cluster. This is done by setting lower values of SHADOW_REC_DLY
on systems that are preferred to handle recovery operations and higher
values of SHADOW_REC_DLY on the remaining systems.
SHADOW_REC_DLY is a dynamic
parameter; its range is 0 to 65535 seconds. The default value is 20
seconds.
For more information about
controlling which systems perform the merge or copy operations, see “Controlling Which Systems Manage Merge and Copy Operations”.
SHADOW_SYS_DISK
A SHADOW_SYS_DISK parameter value of 1 enables shadowing
of the system disk. A value of 0 disables shadowing of the system
disk. A value of 4097 enables a minimerge. The default value is 0.
If you enable a minimerge of the system disk,
you must also configure your system to write a dump to a non-shadowed,
non-system disk of your choice. This is known as dump off system disk
(DOSD). For more information on DOSD, see the HP OpenVMS
System Manager's Manual, Volume 2: Tuning, Monitoring, and Complex
Systems.
In addition, you must specify a system-disk, shadow-set
virtual unit number with the SHADOW_SYS_UNIT system parameter, unless
the desired system disk virtual unit number is DSA0.
SHADOW_SYS_TMO
You can use the SHADOW_SYS_TMO parameter in two ways:
during the booting process and during normal operations. SHADOW_SYS_TMO
is a dynamic parameter that you can change on a running system.
During the booting process, you can use this parameter
on the first node in the cluster to boot and
to create a specific shadow set. If the proposed shadow set is not
currently mounted in the cluster, use this parameter to extend the
time a booting system waits for all former members of the system disk
shadow set to become available.
The second use of this parameter comes into effect
once the system successfully mounts the shadow set and begins normal
operations. Just as the SHADOW_MBR_TMO parameter controls the time
the operating system waits for failing members of an application disk
shadow set to rejoin the shadow set, the SHADOW_SYS_TMO parameter
controls how long the operating system waits for failing members of
a system disk shadow set. All nodes using a particular system disk
shadow set must have their SHADOW_SYS_TMO parameter equal to the same
value, after normal operations begin. Therefore, after booting, this
parameter applies only to members of the system disk shadow set.
The default value is OpenVMS version specific.
You can set a range of up to 65,535 seconds if you want the system
to wait longer than the default for all members to join the shadow
set.
SHADOW_SYS_UNIT
The SHADOW_SYS_UNIT parameter, which must be used when
the SHADOW_SYS_DISK parameter is set to 1, contains the virtual unit
number of the system disk.
The SHADOW_SYS_UNIT parameter is an integer value
that contains the virtual unit number of the system disk. The default
value is 0. The maximum value allowed is 9999. This parameter is effective
only when the SHADOW_SYS_DISK parameter has a value of 1. This parameter
must be set to the same value on all nodes that boot off a particular
system disk shadow set. SHADOW_SYS_UNIT is not a dynamic parameter.
SHADOW_SYS_WAIT
Use the SHADOW_SYS_WAIT parameter to extend the time a
booting system waits for all current members of a mounted system disk
shadow set to become available to this node.
SHADOW_SYS_WAIT is a dynamic parameter that you can change on a running
system (for debugging purposes only). The shadow set must already
be mounted by at least one other cluster node for this parameter to
take effect. The default value is 256 seconds. Change this parameter
to a higher value if you want the system to wait more than the 256-second
default for all members to join the shadow set. This parameter has
a range of 1 through 65,535 seconds.
