Changes to Database Manager configuration parameters in Db2 11.5.1
To support new and changed functionality in the release of Db2 11.5.1, changes were made to some Database Manager configuration parameters.
Attention: Changes in version 11.5.1 affect the following Db2 products:
- The single container deployments of Db2 Warehouse and IBM Integrated Analytics System (IIAS)
- The container micro-service deployment of Db2 on Red Hat OpenShift
- The Db2 cartridge used by IBM Cloud Pak for Data
Details
Resolution
Review the changes to learn which ones might affect your existing applications and to determine whether you have to adjust them.
An internal change has been made to the Fast Communications
Manager (FCM) node failure model. With this new model, the behavior that is equivalent to
that seen in past releases is an extended timeout with no retries.
Note: The previous FCM node
failure model did this multiplication internally and only attempted a single connection instead of
retrying.
To maintain the behavior from previous releases, two Database Manager configuration parameters are now modified upon instance migration. When migrating from Db2 11.5.0.0 and earlier Db2 pureScale to an 11.5.1 and later Db2 pureScale environment, you must translate the conn_elapse and max_connretries default parameters to preserve the behavior.
The following database manager configuration parameters can have a different value after
migration:
To take advantage of the new FCM node failure model, change the values of these two
parameters as appropriate. See the documentation for each parameter to know more.
| Configuration parameters | Change in behavior when migrating to Db2 11.5.1 and later |
|---|---|
| conn_elapse | The value of the parameter is set to ( conn_elapse *
max_connretries + 1 ) Note: The calculation for
conn_elapse uses the value of max_connretries that exists
prior to migration.
|
| max_connretries | After the conn_elapse value is set, max_conn is set to 0. |