admin sqm_process_time

Displays an estimate of the time Replication Server requires to process the remaining data in queues.

Syntax

admin sqm_process_time, 
      {<data_server>, <database>, <q_type> |
      <q_number>, <q_type> |
      <replication_server> |
      <primary_data_server>, <primary_database>, <replicate_data_server>, <replicate_database> |
      <primary_data_server>, <replicate_data_server>}

Parameters

<data_server>

The data server with the database that connects to the path with the specific queue for which you want to estimate the process time

<database>

The name of the database

<q_type>

The type of queue:

0 – outbound queue
1 – inbound queue

<q_number>

The ID number that Replication Server assigns to the queue. The number can be found in the output of the admin who, sqm command.

<replication_server>

The name of the Replication Server hosting the queue for which you want to estimate the process time

<primary_data_server>, <primary_database>, <replicate_data_server>, <replicate_database>

The primary and replicate data servers and databases if you want to estimate the total of the inbound and outbound queue process times

Examples

Example 1

Estimate the time that Replication Server requires to process the remaining data in the inbound queue in the replication path between the pubs1 database in the LDS primary data server and the primary Replicaton Server:

 admin sqm_process_time, LDS, pubs1, 1

You see all the SQM reader modules for the queue you specified and the estimated time to process the backlog of transactions within each reader module rounded up to the nearest second:

Reader                          Estimated Time to Process
------------------------------- -------------------------
104:1 DSI 107 SYDNEY_DS.pubs1                           0
104:1 DIST LDS.pubs1                                    1

In the example, Replication Server estimates that the time to process the backlog of transactions in the inbound queue from the pubs1 database (<q_number>: 104) of the LDS.pubs1 logical connection is 1 second. In non-warm standby environments, the inbound queue only displays one reader. The second reader in this example is for the standby DSI in the warm standby logical connection. Replication Server estimates that 0 seconds is required to process the backlog from the standby DSI.

Example 2

Estimate the time that Replication Server requires to process the remaining data in the outbound queue in the replication path between the database with <q_number> of 104 and Replication Server:

admin sqm_process_time, 104, 0

You see the SQM reader modules for the queue you specified and the estimated time to process the backlog of transactions within each reader module rounded to the nearest second:

Reader                          Estimated Time to Process
------------------------------- -------------------------
104:0  DSI 104 LON_DS.rdb1                              1

Example 3

Estimate the time that Replication Server requires to process the remaining data in the outbound queue in the route to the TOKYO_RRS replicate Replicaton Server:

admin sqm_process_time, TOKYO_RRS

You see all the SQM reader modules for the outboud queue and the estimated time to process the backlog of transactions within each reader module rounded to the nearest second:

Reader                          Estimated Time to Process
------------------------------- -------------------------
TOKYO_RRS                                               1

Example 4

Estimate the time that Replication Server requires to process the remaining data in both the inbound and outbound queues in the route from pdb1 primary database in the NY_DS primary data server to the rdb1 replicate database in the LON_DS replicate data server:

 admin sqm_process_time, NY_DS, pdb1, LON_DS, rdb1

You see a value that is the total for the inbound and outbound queue estimated times to process the backlog of transactions rounded to the nearest second:

Estimated Time to Process
-------------------------
                        3

Usage

There is a one minute sampling period after you execute admin sqm_process_time as Replication Server enables counters to collect data. Wait for the command to display the output , although there appears to be no response to the command during the sampling period. The one minute sampling period occurs only on the initial collection of data from a a specific queue. If the queue has caught up with processing the transactions, the response from admin sqm_process_time returns immediately.
The output of admin sqm_process_time displays all SQM reader modules for the queue and the estimated time to process the backlog of transactions within each SQM reader.
The estimated time is rounded to the nearest second.
Specify only the Replication Server to obtain an estimate of the time to taken to process the remaining data in the outbound queue of the route.
Use admin sqm_process_time with the option to specify both the primary and replicate data servers and databases if you are performing a planned failover and need to an estimate of the total of the time that is going to be taken to process transactions in both the inbound and outbound queues.
Note Both the primary and replicate Replication Servers must be running when you use this option.
You can only use admin sqm_process_time in a replication environment containing a maximum of two Replication Servers with a direct route between them. You cannot use the command in environments with indirect or intermediate routes. If you request an estimate between two data servers, admin sqm_process_time ignores paths that use indirect routes.
To estimate the queue process time for a full path, at least one subscription must exist from the primary to the replicate database. Therefore, admin sqm_process_time does show the estimates for single logical connections but you cannot use admin sqm_process_time to display estimates for warm standby path between the active and standby databases. Similarly, you can use admin sqm_process_time to show the estimate for any one connection in a Multi-Path Replication™ environment but admin sqm_process_time does not show the estimates for all the connections or for an end-to-end primary to replicate path.
If you request a queue process time estimate of the backlog between two data servers and there are no paths between the data servers, you see this error message:
```
No path was found from ‘<primary_data server>’ to ‘<replicate data server>’. 
```
The output may display the max INT value of 2,147,483,647s instead of the estimated process time in seconds if:
- Replication Server has just started,
- The counters did not collect any data within the one minute sampling time, or
- The reader module such as DSI is not active

Permissions

Any user may execute this command.