Abstract: A method for removing stale messages from a storage controller is disclosed. In one embodiment, such a method includes querying, by a host system, a storage controller to determine ownership of a lock on the storage controller. The host system receives, in response to the query, information indicating that the lock has been granted to the host system. This allows the host system to treat the lock as being granted even though the host system has not received a “lock granted” message from the storage controller. Before using the lock, the host system sends, to the storage controller, an instruction to clear any stale messages on the storage controller related to the lock. A corresponding system and computer program product are also disclosed.

Abstract: A method for removing stale messages from a storage controller is disclosed. In one embodiment, such a method includes querying, by a host system, a storage controller to determine ownership of a lock on the storage controller. The host system receives, in response to the query, information indicating that the lock has been granted to the host system. This allows the host system to treat the lock as being granted even though the host system has not received a “lock granted” message from the storage controller. Before using the lock, the host system sends, to the storage controller, an instruction to clear any stale messages on the storage controller related to the lock. A corresponding system and computer program product are also disclosed.

Abstract: A method for identifying a set of records in a database for relocation during an on-line database reorganization (ODBR) process. The relocating includes, for each of the records in the set: requesting a record reorganization lock (RRL) on the record, receiving the RRL, retrieving data from the current location of the record, storing the data in the new location of the record, and releasing the RRL. A request is received to update a target record in the set of records with updated data. The receiving of the request is from an application and overlaps in time with at least a portion of the relocating. Independent of the application, a RRL is received on the target record and the target record is updated by writing target data to the target record in the current location and writing the target data to the target record in the new location.

Abstract: Changes in locking configurations are managed. A locking configuration includes one or more locking facilities and the one or more locking facilities include one or more locks. When a change in a locking configuration, either planned or unplanned, is detected, one or more of the locks are redistributed. For example, one or more of the locks affected by the change are moved from one or more of the locking facilities to one or more other locking facilities. This redistribution is independent of the location of the data associated with the one or more locks being moved. Thus, data and locks may be moved independently of one another and without affecting one another. Further, the redistribution can take place while the system is available or mostly available, thereby minimizing downtime.