Prioritizing Merge and Copy Operations

Default Management of Merge and Copy Operations

If a system fails or if it aborts a shadow set, most commonly through mount verification, it is termed as a significant event. When one of these significant events occurs, all systems in the cluster are notified automatically. This notification causes all shadow server processes to stop any full merge or full copy operations and release all the resources performing these operations. Thus, every system can reallocate its resources to newer, higher-priority work.

After a predetermined delay, each system with a non-zero SHADOW_MAX_COPY setting begins to process the shadow sets that are in a transient state, according to their priority. The predetermined delay is governed by the new system parameter SHADOW_REC_DLY. (For more information about SHADOW_REC_DLY, see Table 3-1 and “Controlling Which Systems Manage Merge and Copy Operations”.) Every system allocates the available SHADOW_MAX_COPY resources based on a shadow set's priority.

A shadow set is in a steady state when it is known that all members contain identical data. If a shadow set has one or more of the following operations pending, or one operation active, it is said to be in a transient state:

While a combination of these transient states is valid, only one operation at a time can be performed. For example, consider that HBMM is not enabled. After a device is added to a shadow set, it is marked as being in a full copy transient state. If the system on which this shadow set is mounted fails, the shadow set is further marked as being in a full merge state. In this example, the full copy operation is performed before the full merge is started.

	NOTE: The priority assigned to a shadow set does not affect the hierarchy of transient state operations.

Hierarchy of Transient State Operations

Shadow set operations for a specific shadow set are performed in the following order:

Assigning Priorities to Shadow Sets

When first mounted on a system, every shadow set is assigned a default priority of 5000. You can assign a unique priority to every mounted shadow set on a per-system basis using the SET SHADOW/PRIORITY=n DSAn command. Every shadow set can have a unique priority per system, or shadow sets can be assigned the same priority. Shadow sets with the same priority are managed in a consistent way for each release. However, the order in which shadow sets with the same priority are managed may change from release to release because of changes to the algorithm. Therefore, if the order is important, assign them different priorities.

The valid range for priority values is 0 through 10,000. The higher the assigned value, the higher the priority. To ensure that high-priority volumes are merged (or copied) before less important volumes, use SET SHADOW/PRIORITY=n DSAncommand to override the default priority assignment on a system.

A priority level of 0 has a unique meaning. It means that the shadow set is not considered for merge or copy operations on this system.

	NOTE: After the notification of a significant event and the allocation of a system's resources, it is not possible to directly affect any of the current merge or copy operations on the system by assigning a different priority level to one or more shadow sets. If you have to re-prioritize one or more shadow sets, you must use another technique, as described in “Managing Transient States in Progress”.

Displaying Shadow Set Priority Values

You can display the priority of a shadow set on a specific system by issuing the following command:

$ SHOW SHADOW/BY_PRIORITY DSAn:   

This command displays the current priority and status of the specified shadow set. If any copy or merge operations are in progress, the node on which the operation is progressing is displayed, along with its progress. For example:

$ SHOW SHADOW DSA1104/BY_PRIORITY 
Device    Mbr                                                        Active
 Name     Cnt  Priority     Virtual Unit State                      on Node
_DSA1104:  2     5000       Merge Active (29%)                         MAX 

You can use the SHOW SHADOW/BY_PRIORITY command to display the priority level and the status for all of the shadow sets that exist on the system. The status indicates whether the shadow set is currently undergoing a copy or merge operation or whether one is required. If either or both operations are underway, the systems on which they occur are identified in the display, as shown in the following example:

$ SHOW SHADOW/BY_PRIORITY 
Device    Mbr                                                        Active
 Name     Cnt  Priority     Virtual Unit State                      on Node
_DSA106:   2    10000       Steady State
_DSA108:   3     8000       Steady State
_DSA110:   3     8000       Steady State
_DSA112:   3     8000       Steady State
_DSA114:   1     7000       Steady State
_DSA116:   1     7000       Steady State
_DSA150:   2     7000       Steady State
_DSA152:         7000       Not Mounted on this node
_DSA154:   3     6000       Steady State
_DSA156:   1     6000       Steady State
_DSA159:   2     5000       Steady State
_DSA74:    3     5000       Merge Active (47%)                       CASSID 
_DSA304:  2+1    5000       Merge Active (30%), Copy Active (3%)     MAX 
_DSA1104:  2     5000       Merge Active (29%)                       MAX 
_DSA300:  2+1    5000       Merge Active (59%), Copy Active (0%)     MAX 
_DSA0:    1+2    5000       Copy Active (83%)                        CASSID 
_DSA3:     2     3000       Steady State
_DSA100:   2     3000       Steady State
_DSA102:   1     3000       Steady State
_DSA104:   3     3000       Steady State
Total of 19 Operational shadow sets; 0 in Mount Verification; 1 not mounted
$

In this example, the 20 shadow sets on this system are displayed in their priority order. In the event of a failure of another system in the cluster that has these shadow sets mounted, the shadow sets are merged in this order on the system.

The Mbr Cnt field shows the number of source members in each shadow set. If members are being added via a copy operation, this is indicated by +1 or +2. Therefore, 2+1 indicates two source members and one member being added. The notation 1+2 indicates one source member and two members being copied into the set.

The summary line provides the total number of shadow sets that were found to be in the various conditions. “Operational shadow sets” are shadow sets that are mounted with one or more members and may or may not have copy or merge operations occurring. These shadow sets are available to applications for reads and writes. “Mount Verification” indicates the number of shadow sets that are in some mount verification state. Shadow sets that have exceeded their mount verification timeout times are also included in this total.

For additional examples, see the SHOW SHADOW examples in the HP OpenVMS DCL Dictionary.

Controlling Which Systems Manage Merge and Copy Operations

When a system fails or aborts a shadow set, this significant event causes every shadow set to be reassessed by all other systems with that shadow set mounted. All active minimerge, full merge, or copy operations cease at this time, returning their resources to those systems. (However, if a system is performing a minicopy operation, that operation continues to completion.)

These systems wait a predetermined amount of time, measured in seconds, before each attempts to manage any shadow set in a transient state. This pause is called a significant-event recovery delay. It is the total of the values specified for two system parameters, SHADOW_REC_DLY and RECNXINTERVAL. (The default value for each is 20 seconds.)

If the value of the significant-event recovery delay is the same on all systems, it is not possible to predict which systems manage which shadow set. However, by making the value of the significant-event recovery delay different on all systems, you can predict when a specific system begins to manage transient-state operations.

Managing Merge Operations

A merge transient state is an event that cannot be predicted. The management of merge activity, on a specific system for multiple shadow sets, can be predicted if the priority level settings for the shadow sets differ.

The following example illustrates how the priority level is used to select shadow sets when only merge operations are involved. In this example:

$ SET SHADOW/PRIORITY=7000 DSA1:
$ SET SHADOW/PRIORITY=3000 DSA42:
! DSA20: and DSA22: are at the default priority level of 5000

When one of the systems in this example fails, all shadow sets are put into a merge-required state. After the significant-event recovery delay time elapses, this system evaluates the shadow sets, and the operations are performed in the following order:

Managing Copy Operations

A copy transient state can be predicted by the user because it is the result of direct user action. Therefore, a full copy operation caused by adding a device to a shadow set is not considered a significant event in the cluster. The copy operation is managed by the first system that has an available resource.

In the following example, assume that there are four shadow sets, and the SHADOW_MAX COPY parameter on this system is equal to 1. Note that the for shadow sets that are not assigned a specific level, a default priority level is assigned.

For the following example, assume that:

The user adds a device to DSA1. This is not a significant event, and this system does not interrupt the full copy operation of the DSA22 in favor of performing the DSA1 full copy operation.

To expand on this example, assume that a system fails (a significant event) before the copy operations have completed. All shadow sets are put into a merge required state. Specifically, DSA1, DSA20, and DSA22 are put into a full merge state, and DSA42 is put into a minimerge state.

After the significant event recovery delay expires, this system begins to evaluate all the shadow sets in a transient state. The operations take place in the following order:

Thus, in this example, the priority level is used to direct the priority of merge and copy operations on this system.

Managing Transient States in Progress

SHADOW_MAX_COPY is a dynamic system parameter that governs the use of system resources by shadowing. Shadowing can be directed to immediately respond to changes in this parameter setting using the following DCL command:

$ SET SHADOW/EVALUATE=RESOURCES 

This command stops all the current merge and copy operations on the system on which it is issued. It then restarts the work using the new value of SHADOW_MAX_COPY.

This command is also useful in other circumstances. For example, if a shadow set has a priority level 0 or another low value, the SET SHADOW /PRIORITY=n command can be used to increase the value. Then, using the /EVALUATE=RESOURCES qualifier, the priority of shadow sets in a transient state is reevaluated.

The /PRIORITY and /EVALUATE=RESOURCES qualifiers can be used on the same command line.

When a significant event occurs, all of the SHADOW_MAX_COPY resources are applied. If the value of SHADOW_MAX_COPY is modified using the SYSGEN SET and WRITE ACTIVE commands, and then a SET SHADOW/EVALUATE=RESOURCES is issued, the new value of SHADOW_MAX_COPY has a direct and immediate affect.

To determine which system is controlling a transient operation, enter the following command:

$ SHOW SHADOW/ACTIVE DSAn:

To determine the priority values assigned to each shadow set, enter the following command:

$ SHOW SHADOW/BY_PRIORITY DSAn: