KR100798997B1 - Restarting Method of Dual Processor - Google Patents

Abstract

The present invention relates to a method of restarting a redundant processor in which the redundant processor of the exchange can resume the standby state by restoring the processor state before the dual down occurs when the dual down occurs during the standby loading process.

If a dual down occurs during standby loading at the exchange, when the redundant processor is restarted, the operating system loaded to each side of the redundant processor may return the previous state information of the processor that is loaded. Since it is not preserved, both active and standby sides of the redundant processor are loaded via absolute path and restarted by the old software package stored on the active side of the redundant disk. There is a problem that must be performed again.

The present invention stores the state of a redundant processor in an external processor state server during a standby loading process by using the state information stored in the processor state server during restart even if a dual down occurs during standby loading. Restoring a suspended standby loading by restoring the processor state before the dual down occurred.

Description

Restarting method of dual processor

1 is a block diagram illustrating a processor and a disk constituting an exchanger.

2 is a block diagram showing a redundant structure of a processor and a disk envisioning an exchanger;

3 is a block diagram illustrating an apparatus for restoring a state before dual down upon restarting a processor according to the present invention.

4 is a flowchart illustrating a process of storing processor state information of each side constituting a redundant processor in a processor state server.

5 is a flowchart illustrating a process of restoring a state prior to dual down to continue standby loading.

6 is a flowchart illustrating a process of transmitting processor state information by a processor state server in response to the state information request transmitted from FIG. 5.

* Description of the symbols for the main parts of the drawings *

40: Disk 50: Processor Status Server

60: redundant processor

60a, 60b: active side and standby side of redundant processors

70: redundant disk                 

70a, 70b: active side and standby side of redundant disk

The present invention relates to a method for restarting a redundant processor. In particular, the redundant processor of the switch exchanges the redundant processor to resume the standby state by restoring the processor state before the dual down occurs when the dual processor occurs during the standby loading process. It relates to a restart method of the processor.

In general, the exchanger is composed of a plurality of main processors (hereinafter, referred to as 'processors') 10 as shown in FIG. 1. When the exchanger starts up the processor 10 constituting the exchanger, the processor ( 10) First, an operating system (OS) 10a is loaded from a disk 12 storing a program necessary for operation, and a program loader (OS) 10 is loaded by the operating system 10a. 10b) is started. The program loader 10b loads and executes a software package for operating the processor 10 from the disk 12 so as to perform a function of the processor 10.

In the exchange, the processor 20 and the disk 30 are both active sides 20a and 30a and standby side 20b as shown in FIG. 2 for stable operation of the system. 30b), and when the program needs to be changed in the operating exchange, the stand-by loading function using the standby side 20b of the redundant processor 20 is used to minimize service interruption. do.

The standby loading is a method of replacing a software package (hereinafter, referred to as 'package') generated during operation, and the active side 20a of the processor 20 having a redundant structure continues to service normally, and the standby side 20b This means that a new package is loaded in the package, and the package is confirmed to be normally operated, and then the package is applied to the active side 20a. This method is designed to minimize service downtime, unlike simply dumping a scene package and rebooting the processor.

A method of changing a package executed in the redundant processor 20 by using the standby loading function will be described in more detail. First, the disk 30 is moved to the active side 30a and the standby side 30b to dump the new package. Redundant to Currently, the old package is stored in both the active / standby side 20a and 20b of the redundant processor 20 and the active / standby side 30a and 30b of the redundant disk 30.

Thereafter, the scene package is dumped to the standby side 30b of the redundant disk 30 and the scene package is loaded to the standby side 20b of the redundant processor 20. Now, the active side 20a, 30a of the redundant processor 20 and the redundant disk 30 both store the old package, and the standby side 20b, 30b of the redundant processor 20 and the redundant disk 30 All new packages are saved.

At this time, the state of the standby side 20b of the redundant processor 20 is switched to the active state, and the state of the active side 20a of the redundant processor 20 is switched to the standby state. Similarly, the state of the standby side 30b of the redundant disk 30 is switched to the active state, and the state of the active side 30a of the redundant disk 30 is switched to the standby state. That is, processor A 20a, which was previously active, is in a standby state, processor B 20b, which was previously in a standby state, is in an active state, and disk a 30a, which was previously in an active state, is in a standby state, The disk b 30b, previously in the standby state, becomes active. Accordingly, the active sides 20b and 30b of the redundant processor 20 and the redundant disk 30 store new packages, and the standby sides 20a and 30a of the redundant processor 20 and the redundant disk 30 are stored. You are saving the old package.

Finally, the old package stored in the standby side 30a of the redundant disk 30 is updated with the new package, and the existing old package executes all data executed in the active side 20b of the redundant processor 20. Dual copies are made to the standby side 20a of the redundant processor 20. At the end of standby loading, the active / standby sides 20a, 20b, 30a, and 30b of the redundant processor 20 and the redundant disk 30 both store new packages.

In each step of performing the above-described standby loading method, the state change of the redundant processor 20 and the redundant disk 30 is shown in Table 1. Here, 'active' or 'standby' indicates an operation state of the redundant redundant processor 20 or the redundant disk 30, and 'OLD' or 'NEW' is stored in the redundant redundant processor 20 or the disk 30. Indicates package status.

State Change of Redundant Processors State Change of Redundant Disk Processor A Processor B Disk a Disk b    Active / OLD Active / OLD Standby / OLD Standby / NEW    STANDBY / OLD STANDBY / NEW ACTIVE / NEW ACTIVE / NEW    Active / OLD Active / OLD Standby / OLD Standby / NEW    STANDBY / OLD STANDBY / NEW ACTIVE / NEW ACTIVE / NEW

However, in the above-described exchange, the standby loading method using the standby side of the redundant processor is a dual down in which the active side and the standby side constituting the redundant processor stop performing their operations during standby loading. If a symptom occurs, when restarting the redundant processor, the operating system loaded on each side of the redundant processor is in the previous state of the processor to which it belongs, such as whether it is active or standby, or whether the loaded package is a new or old package. Since the state information of the redundant processor is not preserved, both the active side and the standby side of the redundant processor are loaded with the old package stored in the active side of the redundant disk through the absolute path and restarted by the operator. The loading procedure must be performed from the beginning again, which not only prevents the stable and reliable operation of the whole system, but also causes a long service downtime to be minimized.

 SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and an object thereof is that even if a dual down occurs during a standby loading process, the redundant processor of the exchange stores the state of the redundant processor in an external processor state server during the standby loading process. When restarting, the state information stored in the processor state server is used to restore the state of the processor before the dual down occurs so that the suspended standby loading can be continued.

In order to achieve the above object, a feature of the present invention includes: transmitting state information of a redundant processor to a processor state server and storing the disk in a disk during standby loading; Requesting state information from one of the redundant processors to the processor state server when restarting the redundant processor after the dual down; If the state information request is a state request first received after the dual down, setting the processor requesting the state information to an active state and storing the state information in the disk; Setting an operating state of another processor constituting the redundant processor to a standby state and storing the operating state in the disk; If the operating state of the configured redundant processor and the operating state of the redundant processor before the dual down are not the same, exchanging and setting states of a software package and storing the same in the disk; Setting the state of the software package to be identical to the state of the software package before the dual down and storing the same on the disk when the operating state of the redundant processor is identical to that of the redundant processor before the dual down; Transmitting status information of the redundant processor stored in the disk; Receiving the state information of the redundant processor to set the current state of the redundant processor and at the same time to continue the standby loading suspended after the dual down to provide a method for restarting the redundant processor.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The apparatus for restoring the state before the dual down upon restarting the processor according to the present invention, as shown in FIG. 3, the redundant processor 60, the redundant disk 70, the processor state server 50 and Made of a disk 40. Here, since the redundant processor 60 and the redundant disk 70 include the functions of the redundant processor 20 and the redundant disk 30 shown in FIG. 2, redundant description of the same functions will be omitted.

Here, the processor state server 50 receives the state information of the processor A 60a and the processor B 60b of the redundant processor 60 which is in the standby loading process, and stores the state information in the disk 40. When the processor A 60a and the processor B 60b of the processor 60 are restarted after the dual down, the processor state information stored in the disk 40 is transmitted to the processor A 60a or the processor B 60b. Do it. In addition, processor A 60a or processor B 60b constituting the redundant processor transmits its changed state information to the processor state server 50 when its state changes during the standby loading process, and restarts after dual down. In this case, the processor state server 50 requests the processor state information to determine its current state, and then continues the standby loading process stopped after the dual down. In this case, the processor state information means active / standby state information which is an operation state of the redundant processor 60 and new / old state information which is a state of a package loaded in the redundant processor 60. .

4 is a flowchart illustrating a process of storing processor state information of each side 60a and 60b constituting the redundant processor 60 in the processor state server 50.

The storing of the processor state information in the processor state server 50 includes a step of transmitting a state information storage request message including the current state information of each processor 60a and 60b constituting the redundant processor 60. And the processor state server 50 receives the state information storage request message and stores the processor state information included therein in the disk 40. Here, it is preferable to transmit and receive a message between the sides 60a and 60b of the redundant processor 60 and the processor state server 50 by the TCP / IP packet exchange method.

In more detail, first, each processor (operating system or program loader) 60a or 60b constituting the redundant processor 60 has its own information in a processor information field of a packet constituting the state information storage request message. The name of the system and the name of the processor to which the system belongs is set (S41), and the status information of the processor is set in the status field of the packet (S42). After calculating and setting the packet length and checksum (S43), setting the message type field value to '1', for example (S44), and configuring the state information storage request message as a packet, the processor state server Transmit to 50 (S45).

Accordingly, the processor state server 50 receives the state information storage request message composed of the packet, analyzes it (S51), first determines whether the message type of the received message is '1' (S52), and If the message type is '1', the processor state information included in the message is extracted and stored in the disk 40 (S53). The disk 40 has a status information table composed of information such as a system name, a processor name, and a processor status as shown in Table 2, for example, and stores the processor status information in the status information table.

System name Processor name Processor status Processor A ISDN exchange OMP ACT / OLD Processor B ISDN exchange OMP SBY / NEW

(ACT: Active state, SBY: Standby state, OLD: Old package, NEW: New package)

Table 2 shows that the operating state of the processor A constituting the redundant OMP of the ISDN exchange is the active state and the old package is loaded, the operating state of the processor B is the standby state and the new package is loaded.

On the other hand, after the processor state information is stored in the disk 40, the state information storage request message is transmitted by configuring a state storage confirmation message in which the message type is set to, for example, '2' for the state information storage request message. It transmits to one processor (S54).

 However, if the message type of the received message is not '1' in step S52, the received message is discarded (S55).

As such, when the processor 60a or 60b receives the confirmation message transmitted from the processor status server 50 with respect to the state information storage request message, the processor 60a or 60b analyzes the confirmation message (S46). 2 ', and if the acknowledgment message type is' 2', analyzing the acknowledgment message to determine whether the checksum is the same as the checksum of the state storage request message, Check whether it matches (S48). As a result of the determination, when the checksums are the same, a response message in which the message type is set to, for example, '3' is configured for the confirmation message and transmitted to the processor status server 50 (S49).

However, when the message type is not '2' in the process of S47 or when the checksum is not the same in the process of S48, the confirmation message is discarded (S50).

FIG. 5 is a flowchart illustrating a process of continuing standby loading by restoring a state before dual down when restarting a processor after dual down.

When a dual down occurs during standby loading and restarts each of the processors 60a or 60b constituting the redundant processor 60, the operating system first transmits the processor status to the processor state server 50 according to the process shown in FIG. After requesting the information (S510), the processor state information is received and the current state of the processor 60a or 60b is set according to the received processor state information (S520). As a result of setting the state of the processor 60a or 60b, it is determined whether the state of the package loaded in the processor 60a or 60b is an old package (S530), and the old package is loaded into the processor 60a or 60b. If there is, the program loader is loaded from the active side 70a of the redundant disk 70 (S540), and if a new package is loaded in the processor 60a or 60b, the standby side of the redundant disk 70 ( The program loader is loaded from 70b) (S550).

When the program loader is loaded on the processor 60a or 60b, the program loader requests processor state information from the processor state server 50 according to the process illustrated in FIG. 4 (S560). Received to set the current state of the processor (60a or 60b) (S570). Thereafter, by analyzing the current state of the set processor (60a or 60b) to identify the standby loading step that proceeded until the dual down occurs, and continues the standby loading from the next step (S580).

FIG. 6 is a flowchart illustrating a process in which the processor state server 50 transmits processor state information to a state information request transmitted by an operating system or a program loader in operation S510 or S560 of FIG. 5.

First, when the processor status server 50 receives a status information request from an operating system or a program loader of the processor 60a or 60b (S610), is the status information request transmitted from an operating system or is transmitted from a program loader. In operation S620, when the status information request is transmitted from the operating system, it is determined whether the status information request is received for the first time after dual down (S630).

In this case, when the state information request is the first state information request received after the dual down, the processor 60a or 60b to which the operating system that has transmitted the state information request belongs is set to an active state (S640) and duplexed. Another processor 60a or 60b constituting the processor 60 is set to a standby state (S650).

Subsequently, it is determined whether the operation state of the newly set processor 60a or 60b in the process of S640 and S650 is the same as the operation state before the dual down stored in the disk 40 (S660). If the operation state is not the same as the operation state before the dual down stored in the disk 40, the state of the package is set to be exchanged with each other (S670). For example, before dual down, the processor A 60a is in an active state and an old package is loaded, and the processor B 60b is in a standby state and a new package is loaded, and state information is stored in the disk 40. When the operating system of the processor B 60b, which is in the standby state, first sends a status information request, the processor state server 50 sets the processor B 60b to the active state, and the processor A 60a is set to the standby state. Here, the operating state of the processor set by the processor state server 50 is opposite to the operating state of the processor stored in the disk 40. In this case, the package state of the processor A 60a and the package state of the processor B 60b are exchanged, and the package state of the processor A 60a set to the standby state is set to the new package, and the processor B (set to the active state ( The package state of 60b) is set to the old package and stored. Here, as a result, the state of the processor A (60a) and the state of the processor B (60b) is set opposite to the state before the dual down, but in the entire redundant processor 60, it is understood that the state information before the dual down is restored as it is. Can be. The said example is put together in Table 3 and shown.

Status information stored on disk (status information before dual down) Newly set status information (status information after dual down) Processor A ACT / OLD SBY / NEW Processor B SBY / NEW ACT / OLD

(ACT: Active state, SBY: Standby state, OLD: Old package, NEW: New package)

As such, when the operating state and the package state of the processor are set, the state of the processor is transmitted to the operating system that has transmitted the state information request (S680).

However, if the operation state of the newly set processor in the process of S660 is the same as the operation state before the dual down stored in the disk 40, the package stored in the disk 40 without exchanging the package state is stored. After setting the state to the current package state as it is, the process of S680 for transmitting the current processor state information is performed. For example, if the processor A 60a is active and the old package is loaded before the dual down, and the processor B 60b is in the standby state and the new package is loaded, the state information is stored on the disk. If the operating system of the processor A (60a) first sent a status information request, the processor state server 50 sets the state of the processor A (60a) to the active state and the package state to the old package as before. On the other hand, the operation state of the processor B 60b is set to the standby state as before, and the package state is set to the new package. In this case, it can be seen that the newly set processor state information is restored to the same state information before the dual down. The said example is put together in Table 4 and shown.

Status information stored on disk (status information before dual down) Newly set status information (status information after dual down) Processor A ACT / OLD ACT / OLD Processor B SBY / NEW SBY / NEW

(ACT: Active state, SBY: Standby state, OLD: Old package, NEW: New package)

Meanwhile, the state information request received by the processor state server 50 in the process of S620 is transmitted by the program loader or the state information request received by the processor state server 50 in the process of S630 is not first. In the case of the received state information request, the processor 110 transmits the processor state information stored in the disk 40 newly set after the dual down to the corresponding program loader or the corresponding operating system (S670).

Although the present invention has been described in detail only with respect to the above-described embodiments, it will be apparent to those skilled in the art that modifications and variations can be made within the spirit and scope of the present invention, and such modifications or changes are within the scope of the claims of the present invention. Will belong.

As described above, according to the present invention, if the state of the redundant processor changes during the standby loading process, the redundant processor transmits it to an external processor state server and stores the data on the disk so that the processor can be stored even after the dual down occurs during standby loading. Upon restarting, interrupted standby loading can be continued by restoring the processor state before the dual down occurs by using the state information stored in the processor state server.

Claims (5)

Transmitting state information of the redundant processor to a processor state server and storing the state information of the redundant processor on a disk during a standby loading process; Requesting state information from one of the redundant processors to the processor state server when restarting the redundant processor after the dual down; If the state information request is a state request first received after the dual down, setting the processor requesting the state information to an active state and storing the state information in the disk; Setting a state of another processor constituting the redundant processor to a standby state and storing the state in the disk; If the operating state of the configured redundant processor and the operating state of the redundant processor before the dual down are not the same, exchanging and setting states of a software package and storing the same in the disk; Setting the state of the software package to be identical to the state of the software package before the dual down and storing the same on the disk when the operating state of the redundant processor is identical to that of the redundant processor before the dual down; Transmitting status information of the redundant processor stored in the disk; Receiving the state information of the redundant processor to set a current state of the redundant processor and continuing the standby loading suspended after the dual down; and restarting the redundant processor. The method of claim 1, Saving the state of the redundant processor, Constructing and transmitting a state information storage request message including a name of a system to which the redundant processor belongs, a name of the redundant processor, and state information of the redundant processor, to the processor state server; And the processor state server storing state information of the redundant processor in a state information table inside a disk. The method of claim 1, If the status information request is not the first status information request received after the dual down, transmitting the status information of the redundant processor stored in the disk, the redundant processor stored in the disk without setting the status of the redundant processor And restarting the redundant processor. The method of claim 1, Setting the state of the redundant processor and transmitting state information according to the state information request, When the status information request is a status information request transmitted from the program loader of the redundant processor, the processor status server transmits the status information of the redundant processor stored in the disk to the program loader. Way. The method of claim 1, Setting the current state of the redundant processor and continuing standby loading at the same time may include: Setting a state of the processor according to the state information received by the operating system of the redundant processor, and then determining whether the current software package state of the redundant processor is an old package; Loading a program loader from the active side of the redundant disk when the current software package state is an old package; Loading a program loader from a standby side of the redundant disk when the current software package state is a new package; And setting a state of the processor according to the state information received by the program loader, analyzing a state of the processor to determine a standby loading progress state, and continuing the standby loading.

Images