Virtual Storage Array Mobility via Locator-Identifier Separation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data center reconfigurations, particularly storage resource migrations, often require significant downtime and service disruptions due to the need for IP address changes and remounting of storage shares, which is cumbersome and disruptive, especially in environments with numerous virtual machines.
Innovation Solution
Implementing a Locator-Identifier Separation Protocol (LISP) to manage endpoint identifiers and routing locators, allowing storage mobility within and between data centers without changing IP addresses, thereby reducing downtime and maintaining service continuity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If storage resources are migrated using existing techniques, then storage mobility is achieved, but service downtime and disruption increase
Solution Approach 1:
The patent introduces a locator-identifier separation protocol where the endpoint identifier (EID) serves as an intermediary between the storage client and the physical storage resource. The EID remains constant while the routing locator (RLOC) changes during migration, allowing seamless redirection without service disruption. This mediator approach resolves the contradiction by enabling storage mobility while maintaining service continuity through the decoupling of identification and location.
Solution Approach 2:
The patent segments the traditional unified address into two separate components: endpoint identifier (EID) and routing locator (RLOC). The EID identifies the storage resource logically, while the RLOC provides physical routing information. This segmentation allows the EID to remain stable for service continuity while the RLOC can be updated for storage mobility, thus resolving the contradiction between mobility and reliability.
2Adaptability or versatility
If IP addresses are changed during storage migration, then storage mobility is enabled, but downtime increases
Solution Approach 1:
The endpoint identifier acts as an intermediary that absorbs the change in physical location (RLOC) without requiring changes to the client-facing address. During migration, the EID remains constant while the RLOC is updated in the mapping database, eliminating downtime associated with IP address changes and remounting operations.
Solution Approach 2:
The system performs preliminary updates to the locator mapping before the actual storage migration completes. The mapping database is updated with the new RLOC associated with the existing EID, so when the migration finishes, clients can immediately access the migrated storage without any address changes or remounting, thus eliminating downtime.
3Adaptability or versatility
If remounting of storage shares is required during migration, then storage mobility is achieved, but operation complexity increases
Solution Approach 1:
The endpoint identifier serves as a stable intermediary that clients continue to use throughout the migration process. Since the EID does not change, clients do not need to remount storage shares or reconfigure their connections. The complexity of migration is handled transparently in the background through RLOC updates in the mapping database, greatly simplifying the operation.
4Productivity
If multiple virtual machines are migrated simultaneously, then productivity is improved, but service disruption increases
Solution Approach 1:
Each virtual machine maintains its own endpoint identifier as a stable intermediary during migration. The locator-identifier separation protocol allows multiple EIDs to be updated independently in the mapping database, enabling simultaneous migration of multiple VMs without mutual interference or cascading disruptions, thus improving productivity while maintaining service continuity.
Data Source
AI summary
Example embodiments of the present invention relate to a method, a system, and a computer program product for implementing a locator-identifier separation protocol for a virtual storage array. The method includes associating an endpoint identifier to a virtual storage array in a datacenter and associating a routing locator with the endpoint identifier in accordance with a locator-identifier separation protocol. Communications with the virtual storage array then may be directed according to the endpoint identifier via its association with the routing locator.


