Multi-path mechanism of storage controller

Source: Internet
Author: User

Multi-path mechanism of storage controller

What is ALUA multi-path mechanism?

ALUA is short for "asypolicric Logical UnitAccess (asynchronous Logical unit access)". It is one of the multi-path mechanisms of front-end controllers. The multi-path mechanism of the front-end Controller determines the read/write performance and reliability of the storage to a certain extent. The existing multi-path mechanism of the front-end controller can be divided into three categories:

A/A: distributed Ric Active/Acivie. For A specific LUN, the target ports of the two storage controllers are active/optimized. The two controllers implement high-speed communication. One IO is sent to the Controller, and the two controllers can participate in the processing at the same time. When one controller is busy, the system automatically implements Load Balancing without the involvement of Server Load balancer software on the host.

ALUA: asypolicric Active/Active. For a specific LUN, the target port of a controller is in the active/optimized status in its path, the target port of another controller is in the active/unoptimized status. At A certain time point, a lun only belongs to A certain controller. To achieve load balancing on both sides, you need to throw task A to Controller A and Task B to Controller B, for the same task, only one controller is under control at any time.

A/P: Active/Passive. For A specific LUN, the target port of A controller is in the active/optimized status in its path, the target port of another controller is in the standby status. Its Load Balancing and task processing methods are similar to those of ALUA.

Active/optimized, Active/unoptimized, Standby, and Unavailable are the four access statuses of the target port. In the corresponding access status, the device server (that is, the array Controller) only response to the appropriate command standard (which is determined by the ISO/IEC 14776-453 file Part 453: SPC-3) determines whether logical units can be accessed through a specific destination port at a specific time point.

The status of the target port can be changed. The transition from one status to another is called transition.

Active/optimized: The target port has the ability to access the logical unit immediately.

Active/unoptimized: only the corresponding command standards can be responded, and can be transitioned to Active/optimized.

Standby: it can only respond to the corresponding command standards and can be transitioned to Active/optimized.

Unavailable: You can only respond to a limited command set, but cannot transition to the other three States.

In arrays A/A, the Administrator does not need to specify the default owner of each LUN. When the path fails, the Administrator redirects the Offline Fault path and the IO to another path. During IO redirection, the storage controller takes full consideration of factors such as load balancing and selects the most appropriate path. For applications, the path switching process is transparent and there is almost no latency (the latency is generally several seconds ).

In the ALUA or A/P array, the administrator needs to specify the default owner of each LUN, set the default owner of some Luns to Controller A, and the default owner of some Luns to Controller B, server Load balancer is manually performed between two controllers. If the path fails, I/O traffic will be re-allocated to other available paths, and I/O on the fault path will be stopped. For applications, the path switching process is transparent, but there is a delay (the delay is generally dozens of seconds ).

When developing Load Balancing policies, you must take into account the features of the multi-path software and the multi-path mechanism of the storage array (A/A-ALUA-A/P ).

Theoretically, the most intuitive manifestation of different multi-path mechanisms is that the latency of path switching is different. The A/A mechanism is optimal and almost no delay, while the ALUA mechanism is slightly less efficient and has A delay, the A/P mechanism is the worst and the latency is greater.

Note:

1. ISO/IEC: International Standardization Organization and International Electrotechnical Commission

2. SPC-3: SCSI Primary commands-3

References:

Http://technet.microsoft.com/zh-cn/library/cc725907 (WS.10). aspx

2. http://webstore.iec.ch/preview/info_isoiec14776-453%7Bed1.0%7Den.pdf

3. http://dealer.yesky.com/server/7/8752007.shtml

Contact Us

The content source of this page is from Internet, which doesn't represent Alibaba Cloud's opinion; products and services mentioned on that page don't have any relationship with Alibaba Cloud. If the content of the page makes you feel confusing, please write us an email, we will handle the problem within 5 days after receiving your email.

If you find any instances of plagiarism from the community, please send an email to: info-contact@alibabacloud.com and provide relevant evidence. A staff member will contact you within 5 working days.

A Free Trial That Lets You Build Big!

Start building with 50+ products and up to 12 months usage for Elastic Compute Service

  • Sales Support

    1 on 1 presale consultation

  • After-Sales Support

    24/7 Technical Support 6 Free Tickets per Quarter Faster Response

  • Alibaba Cloud offers highly flexible support services tailored to meet your exact needs.