CN107070699A - Controller monitoring is managed in storage system redundancy design method and device - Google Patents

Abstract

The invention belongs to the technical field of storage system monitoring management, disclose the redundancy design method that controller monitoring is managed in a kind of storage system, different network interface cards are connected to by NCSI dual links including the baseboard management controller on controller mainboard, baseboard management controller transfers information to external management elements by NCSI dual links and network interface card, the NCSI double-strand road networks supported using A A or A S redundancy strategies, running substrate Management Controller.The invention also discloses the Redundancy Design device that controller monitoring in storage system is managed, including NCSI dual links sharing module, information transfer module and monitoring management strategy operation module.The present invention is shared by NCSI dual links, when network interface card or external management elements failure, improves the reliability of controller monitoring management.

Description

Redundancy design method and device for controller monitoring and management in storage system

technical field

The invention belongs to the technical field of storage system monitoring and management, in particular to a redundant design method and device for controller monitoring and management in a storage system.

Background technique

In the current data center deployment, disk arrays are widely used. In order to improve the high availability of storage devices, the number of controllers on the data link has been developed from dual controllers to multiple controllers, and a redundant design has been implemented on the entire path. On the other hand, humanized and intelligent management of massive data has become a major demand, and has increasingly become a hot spot in storage management. In the face of increasing system reliability requirements, users not only expect full redundancy of data paths, but also hope that monitoring and management can also achieve redundant design and hot replacement.

In the past, general-purpose disk array systems were integrated with SAS expander, but more dual-controller disk array head products have appeared recently. More emphasis is placed on the scalability of processors, caches, and interfaces on the controller, and the SAS expander with SES management functions is no longer integrated. At the same time, this kind of high-end multi-control storage system in the industry mostly adopts a modular design, and each module in the system can be hot-swapped for maintenance. Higher requirements are placed on the design of the system. The monitoring and management of such systems cannot use traditional SES, and cannot directly use a single BMC chip to provide monitoring and management services like a server.

Contents of the invention

The purpose of the present invention is to provide a redundant design method and device for controller monitoring and management in a storage system. The present invention uses NCSI dual-link sharing, and when the network card or external management unit fails, the reliability of controller monitoring and management is improved. .

In order to achieve the above object, the present invention adopts the following technical solutions:

A redundant design method for controller monitoring and management in a storage system, comprising the following steps:

The baseboard management controller on the controller motherboard is connected to different network cards through NCSI dual links;

The baseboard management controller transmits information to the external management unit through the NCSI dual link and the network card;

Adopt the A-A or A-S redundancy strategy, and run the NCSI dual-link network supported by the baseboard management controller.

Preferably, before the baseboard management controller on the controller mainboard is connected to different network cards through NCSI dual links, the method further includes: the baseboard management controller on the controller mainboard collects and monitors information on the boards.

Preferably, after the baseboard management controller transmits the information to the external management unit through the NCSI dual link and the network card, the method further includes: using the dual network port corresponding to the NCSI dual link and the external management unit as an isolated network.

Preferably, the IP address of the internal monitoring and management independent network of the storage is initially generated according to the controller ID and the external management unit ID and then solidified.

Preferably, the A-A redundancy strategy specifically includes: when the network card fails or one of the external management units fails, the IP addresses of the two independent network segments of the baseboard management controller transmit information to the outside at the same time.

Preferably, the A-S redundancy strategy specifically includes: when the network card fails or one of the external management units fails, the two NCSI links connected to the baseboard management controller adopt a floating IP drift strategy, and the floating IP drift strategy That is, when the baseboard management controller obtains the external status notified by the system, it drifts the IP to the correct link for monitoring and management.

The present invention also provides a redundant design device for controller monitoring and management in a storage system, including:

NCSI dual-link sharing module, used to connect the baseboard management controller on the controller motherboard to different network cards through NCSI dual-link;

The information transmission module is used for the baseboard management controller to transmit information to the external management unit through the NCSI dual link and the network card;

The monitoring management strategy operation module is used to adopt the A-A or A-S redundancy strategy to run the NCSI dual-link network supported by the baseboard management controller.

Preferably, it also includes: an information collection and monitoring module, used for the baseboard management controller on the controller main board to collect and monitor the information on the board.

Preferably, it also includes: an isolated network setting module, configured to use the dual network port corresponding to the NCSI dual link and the external management unit as an isolated network.

Preferably, the A-A redundancy strategy specifically includes: when the network card fails or one of the external management units fails, the IP addresses of the two independent network segments of the baseboard management controller transmit information to the outside at the same time; the A-S redundancy strategy The strategy specifically includes: when the network card fails or one of the external management units fails, the two NCSI links connected to the baseboard management controller adopt the floating IP drifting strategy. The floating IP drifting strategy means that the baseboard management controller obtains the system When notifying the external state, the IP drifts to the correct link for monitoring and management.

Compared with the prior art, the present invention has the following advantages:

1. The method of the present invention is different from the previous storage array using SAS expander and SES monitoring protocol. In the hardware part, the baseboard management controller on the controller motherboard is connected to different network cards through NCSI dual links, and the baseboard management controller is connected to different network cards through NCSI dual links. The link and network card transmit information to the external management unit. In the software part, the A-A or A-S redundancy strategy is adopted to run the NCSI dual-link network supported by the baseboard management controller, thereby realizing the redundant monitoring and management of the controller through the combination of software and hardware. The design can effectively improve the usability and practicability of controller monitoring and management in the entire storage system, and improve the available and maintainable level of the storage system. The method of the present invention is different from the single-channel NCSI sharing commonly used in servers, and realizes the management port. redundancy.

2. The dual network port corresponding to the NCSI dual link of the present invention and the external management unit are used as an isolated network, which is used as an independent network for internal monitoring and management of the storage system, and other networks are used to provide external networks for user operation or monitoring system status, so as to realize user The isolation of monitoring and system monitoring network improves the reliability of the system.

3. Adopt the A-A or A-S redundancy strategy and run the NCSI dual-link network supported by the baseboard management controller to ensure that the system monitoring service is not interrupted when the network card or the external management unit fails, which improves the monitoring and management of the controller in the storage system. reliability.

Description of drawings

Fig. 1 is a schematic flow chart of a redundant design method for controller monitoring and management in a storage system of the present invention;

Fig. 2 is a logical block diagram of NCSI dual-link sharing;

FIG. 3 is a schematic structural diagram of a redundant design device for controller monitoring and management in the storage system of the present invention.

detailed description

For ease of understanding, the following explanations are made to some nouns appearing in the present invention:

BMC: Baseboard Management Controller (Baseboard Management Controller), in the server system, we usually use BMC to monitor and manage the health status of the motherboard. Some important parameters on the motherboard such as voltage, temperature, power consumption, etc. are all passed through the BMC monitoring records.

NCSI: NCSI (Network Controller Sideband Interface) is an industry standard defined by the Distributed Management Task Force (DMTF) to support server out-of-band management sideband interface network controllers. A management controller and multiple network controllers. DMTF has defined complete Ethernet-based control command request and response standards for NCSI. In addition, NCSI also has mechanisms such as single thread and timeout retransmission.

Below in conjunction with accompanying drawing and embodiment, the specific embodiment of the present invention is described in further detail:

This embodiment provides a redundant design method for controller monitoring and management in a storage system, including:

The baseboard management controller on the controller motherboard is connected to different network cards through NCSI dual links;

The baseboard management controller transmits information to the external management unit through the NCSI dual link and the network card;

Adopt the A-A or A-S redundancy strategy, and run the NCSI dual-link network supported by the baseboard management controller.

Please refer to FIG. 1 and FIG. 2. FIG. 1 is a schematic flowchart of a redundant design method for controller monitoring and management in a storage system of the present invention; FIG. 2 is a logical block diagram of NCSI dual-link sharing; this embodiment provides a storage system A redundant design method for controller monitoring and management, comprising the following steps:

Step S101, the baseboard management controller on the controller mainboard collects and monitors information on the board;

The BMCs on the motherboards of controller 0 and controller 1 collect and monitor information on the boards.

Step S102, the baseboard management controller on the controller mainboard is connected to different network cards through NCSI dual links, and the network cards can be independent network cards or system CPU sharelink network cards;

The BMCs on the motherboards of controller 0 and controller 1 are respectively connected to NIC0 and NIC1 through NCSI dual links.

Step S103, the baseboard management controller transmits the information to the external management unit through the NCSI dual link and the network card;

The BMC on the motherboard of controller 0 and controller 1 transmits information to external management unit A through an NCSI link and NIC0, and the BMC transmits information to external management unit B through another NCSI link and NIC1.

Step S104, using the dual network port corresponding to the NCSI dual link and the external management unit as an isolated network, used as an independent network for internal monitoring and management of the storage system, and using other networks as an external network for providing user operations or monitoring system status, realizing user The isolation of the monitoring and system monitoring network improves the reliability of the system; due to the isolation from the user access network, the IP address of the internal monitoring and management independent network of the storage is initialized and generated according to the controller ID and the external management unit ID and then solidified, which is convenient for the storage system deployment to improve availability and reliability.

Step S105, using the A-A or A-S redundancy strategy to run the NCSI dual-link network supported by the baseboard management controller.

The A-A redundancy strategy specifically includes that when the network card fails or one of the external management units fails, the IP addresses of the two independent network segments of the baseboard management controller transmit information to the outside at the same time. The A-S redundancy strategy specifically includes that when the network card fails or one of the external management units fails, the two NCSI links connected to the baseboard management controller adopt a floating IP drifting strategy, and the floating IP drifting strategy is the baseboard management When the controller obtains the external status notified by the system, it will drift the IP to the correct link for monitoring and management. The policy switching in the storage system is controlled by the upper-layer storage software, and the controller center can generate and control policies through various information collected on serial ports, networks, and local and remote independent units.

The present invention can effectively improve the usability and reliability of controller monitoring and management in the whole storage system through the design of NCSI double-link sharing and redundancy strategy.

This embodiment also provides a redundant design device for controller monitoring and management in a storage system, including:

NCSI dual-link sharing module, used to connect the baseboard management controller on the controller motherboard to different network cards through NCSI dual-link;

The information transmission module is used for the baseboard management controller to transmit information to the external management unit through the NCSI dual link and the network card;

The monitoring management strategy operation module is used to adopt the A-A or A-S redundancy strategy to run the NCSI dual-link network supported by the baseboard management controller.

Please refer to FIG. 3. FIG. 3 is a schematic structural diagram of a redundant design device for controller monitoring and management in a storage system according to the present invention; this embodiment provides a redundant design device for controller monitoring and management in a storage system, including: information collection and monitoring Module 301, NCSI dual-link sharing module 302, information transmission module 303, isolated network setting module 304 and monitoring management policy operation module 305, described information collection and monitoring module 301 and NCSI dual-link sharing module 302, information transmission module in sequence 303. The isolated network setting module 304 is connected to the monitoring management policy running module 305.

The information collection and monitoring module 301 is used for the baseboard management controller on the controller main board to collect and monitor the information on the board;

The NCSI dual-link sharing module 302 is used to connect the baseboard management controller on the controller motherboard to different network cards through the NCSI dual-link;

The information transmission module 303 is used for the baseboard management controller to transmit information to the external management unit through the NCSI dual link and the network card;

The isolated network setting module 304 is used to use the dual network port corresponding to the NCSI dual link and the external management unit as an isolated network;

The monitoring management policy operation module 305 is used to adopt the A-A or A-S redundancy strategy to run the NCSI dual-link network supported by the baseboard management controller, wherein the A-A redundancy strategy specifically includes: when the network card fails or the external management unit When one of the external management units fails, the IP addresses of the two independent network segments of the baseboard management controller transmit information to the outside at the same time. The A-S redundancy strategy specifically includes: when the network card fails or one of the external management units fails, communicate The two NCSI links connected to the baseboard management controller adopt the floating IP drifting strategy. The floating IP drifting strategy means that when the baseboard management controller obtains the external status notified by the system, it will drift the IP to the correct link for monitoring and management.

What is shown above is only a preferred embodiment of the present invention. It should be pointed out that for those of ordinary skill in the art, some improvements and modifications can also be made without departing from the principles of the present invention. It should be regarded as the protection scope of the present invention.

Claims (10)

1. the redundancy design method that controller monitoring is managed in a kind of storage system, it is characterised in that comprise the following steps：

Baseboard management controller on controller mainboard is connected to different network interface cards by NCSI dual links；

Baseboard management controller transfers information to external management elements by NCSI dual links and network interface card；

The NCSI double-strand road networks supported using A-A or A-S redundancy strategies, running substrate Management Controller.

2. the redundancy design method that controller monitoring is managed in storage system according to claim 1, it is characterised in that Baseboard management controller on controller mainboard is connected to by NCSI dual links before different network interface cards, in addition to：Controller The information on board is collected and monitored to baseboard management controller on mainboard.

3. the redundancy design method that controller monitoring is managed in storage system according to claim 1, it is characterised in that Baseboard management controller is transferred information to after external management elements by NCSI dual links and network interface card, in addition to：By NCSI The corresponding dual network ports of dual link are used as isolation network with external management elements.

4. the redundancy design method that controller monitoring is managed in storage system according to claim 3, it is characterised in that institute The IP address for stating the monitoring management separate network inside storage initializes generation according to controller ID and external management elements ID Solidify afterwards.

5. the redundancy design method that controller monitoring is managed in storage system according to claim 1, it is characterised in that institute A-A redundancy strategies are stated to specifically include：When some failure in network interface card failure or external management elements, substrate Two independent network segment IP address of Management Controller externally transmit information simultaneously.

6. the redundancy design method that controller monitoring is managed in storage system according to claim 1, it is characterised in that institute A-S redundancy strategies are stated to specifically include：When some failure in network interface card failure or external management elements, with base Two NCSI links of board management controller connection are using Floating IP address drift strategy, and Floating IP address drift strategy is substrate management control When device gets the external status of notifications, IP is drifted into correct link and is monitored management.

7. the Redundancy Design device that controller monitoring is managed in a kind of storage system, it is characterised in that including：

NCSI dual link sharing modules, the baseboard management controller on controller mainboard is connected to not by NCSI dual links Same network interface card；

Information transfer module, external management list is transferred information to for baseboard management controller by NCSI dual links and network interface card Member；

Monitoring management strategy runs module, for using A-A or A-S redundancy strategies, what running substrate Management Controller was supported NCSI double-strand road networks.

8. the Redundancy Design device that controller monitoring is managed in storage system according to claim 7, it is characterised in that also Including：

The information on board is collected and monitored to information monitoring module, the baseboard management controller on controller mainboard.

9. the Redundancy Design device that controller monitoring is managed in storage system according to claim 7, it is characterised in that also Including：

Isolation network setup module, for regarding the corresponding dual network ports of NCSI dual links and external management elements as separation net Network.

10. the Redundancy Design device that controller monitoring is managed in storage system according to claim 7, it is characterised in that

The A-A redundancy strategies are specifically included：When some failure in network interface card failure or external management elements When, two independent network segment IP address of baseboard management controller externally transmit information simultaneously；

The A-S redundancy strategies are specifically included：When some failure in network interface card failure or external management elements When, two NCSI links being connected with baseboard management controller are using Floating IP address drift strategy, and Floating IP address drift strategy is substrate When Management Controller gets the external status of notifications, IP is drifted into correct link and is monitored management.