Storage Controller Coredump Preservation During Switchover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In storage networks, coredump processes for diagnostic purposes can be disrupted during switchover operations due to the lack of knowledge about the storage device used by the coredump process, leading to potential loss of diagnostic data and disruptive client access.
Innovation Solution
A first storage controller stores a storage device identifier in a disk mailbox during the coredump process, allowing a second storage controller to perform a switchover operation excluding the storage device used for coredump, thus preserving the coredump data and ensuring continuous client access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the second storage controller performs a switchover operation to take ownership of storage devices from the first storage controller, then client access to data is maintained, but the coredump process is aborted and diagnostic data is lost
Solution Approach 1:
The first storage controller stores the identifier of the storage device being used for the coredump process into a disk mailbox before the switchover operation occurs. This preliminary action allows the second storage controller to identify and exclude the coredump storage device from the switchover, preserving the diagnostic data while still allowing switchover of other storage devices to maintain client access continuity.
2Loss of information
If the switchover operation is delayed to allow the coredump process to complete, then the coredump data is preserved, but client access to data is disrupted
Solution Approach 1:
The storage devices are segmented into two categories: the coredump storage device (identified by the stored identifier) and other storage devices. The switchover operation is applied selectively - excluding the coredump storage device to preserve diagnostic data while including other storage devices to maintain client access continuity. This segmentation allows simultaneous achievement of both goals without requiring a complete delay or complete exclusion.
3Loss of information
If the first storage controller sends the storage device identifier to the second storage controller, then the second storage controller can exclude the coredump device from switchover, but the communication may fail due to network unavailability or software failure
Solution Approach 1:
A disk mailbox serves as an intermediary storage mechanism between the first storage controller and the second storage controller. The first storage controller writes the coredump storage device identifier into this mailbox, which the second storage controller then reads. This intermediary approach bypasses the need for direct communication between the controllers, ensuring the identifier is reliably transmitted even when network or software communication fails during the failure event.
Data Source
AI summary
One or more techniques and/or computing devices are provided for preserving coredump data. A first storage controller, of a first storage cluster, may have a disaster recovery relationship with a second storage controller of a second storage cluster. When the first storage controller fails, the first storage controller performs a coredump process to dump memory contents of the first storage controller into a storage device. During implementation of the coredump process, the first storage controller stores a storage device identifier of the storage device into a disk mailbox. Upon detecting the failure, the second storage controller reads the storage device identifier from the disk mailbox. The second storage controller performs a switchover operation to change ownership of storage devices, but excluding the storage device used by the coredump process, from the first storage controller to the second storage controller for providing clients with failover access to the storage devices.