Shared Memory Failover for Logical Partitions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In storage apparatus capable of executing multiple logical partitions (LPARs), swift failover to another LPAR is necessary when a fault occurs in one LPAR, as existing technologies do not efficiently manage the transition of data input/output processes.
Innovation Solution
A virtualization mechanism is implemented that provides a shared memory area accessible by both LPARs, where the first LPAR stores takeover information, enabling the second LPAR to acquire and execute the data input/output processes upon detecting a fault in the first LPAR.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a fault occurs in one LPAR and failover is executed to another LPAR, then service continuity is maintained, but the time required for failover execution is prolonged due to complex data transfer and process takeover procedures
Solution Approach 1:
The patent applies preliminary action by pre-preparing takeover information in a shared memory area before faults occur. The second LPAR预先 acquires necessary data and process information from the first LPAR through the shared memory, so that when a fault occurs, the failover can be executed immediately without time-consuming data transfer procedures.
2Reliability
If existing failover technologies are used, then fault tolerance is achieved, but the complexity of managing data input/output process transitions between LPARs increases
Solution Approach 1:
The patent introduces a shared memory area as an intermediary between LPARs. This shared memory serves as a mediator that automatically facilitates data and process information exchange during failover, eliminating the need for complex manual management of data input/output process transitions while ensuring fault tolerance.
3Speed
If swift failover is implemented through direct process takeover, then failover speed is improved, but the risk of data loss or inconsistency during the transition increases
Solution Approach 1:
The patent applies copying by creating a copy of the necessary takeover information in the shared memory area before failover execution. The second LPAR acquires a copy of the data and process information from the first LPAR, allowing swift direct process takeover while ensuring data consistency through the use of this pre-copied information.
Data Source
AI summary
A storage apparatus is provided with a virtualization mechanism which manages first and second LPARs (logical partitions) assigned respectively with first and second logical resources acquired by logically partitioning physical resources in the storage apparatus. The virtualization mechanism provides a shared memory area which is an area based on a memory and can be accessed by the first and second LPARs. The first LPAR stores information required for taking over a data input/output process handled by the first LPAR to the second LPAR, in the shared memory area. When detecting that a fault occurs in the first LPAR, the second LPAR acquires the information required for taking over from the shared memory area and takes over and executes the data input/output process formerly handled by the first LPAR on the basis of the information required for taking over.