Electrical computers and digital processing systems: memory – Storage accessing and control – Control technique
Reexamination Certificate
2001-11-16
2003-08-26
Nguyen, Hiep T. (Department: 2187)
Electrical computers and digital processing systems: memory
Storage accessing and control
Control technique
C711S165000
Reexamination Certificate
active
06611903
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a storage subsystem and its controlling method, and more particularly to a technique which is effective by being applied to such as the technique of multiplexing data by data copy among a plurality of storage subsystems installed at remote places or the like.
Due to the progress of information society in recent years, there has been an explosive increase in the amounts of data stored in storage devices of computer systems used in information processing in various fields of social activities, and there has been a demand for high reliability for both the data and the system. As a technique for satisfying both the reliability of data and the assurance of the system operation, the multiplexing of storage devices themselves is being undertaken. The data outputted from a host computer is copies not only to the storage device directly connected thereto but also to other storage devices via that storage device.
Generally, this technique is called remote copy, and its application to computer systems requiring high reliability is being implemented. According to this remote copy technique, even if a fault has occurred in one storage device and the storage device has lapsed into an inoperable state, the system can be operated by using the data of another storage device.
A remote-copy controlling system has been realized in which when a host computer stores data in a storage device connected to the host computer (hereafter, this storage device will be referred to as a master-side storage device), the master-side storage device is able to, synchronously or asynchronously with the data transmission, copy the data to another storage device connected to the master-side storage device (hereafter, this storage device will be referred to as a remote-side storage device).
In the event that a fault has occurred in the master-side storage device and the master-side storage device has lapsed into an inoperable state, the data remaining in the remote-side storage device can be used as they are by merely changing over the storage device which becomes active to the remote side.
In addition, remote copy to remote plates is also being practiced via public lines. Even if a large-scale natural disaster or the like has occurred in the place or the city where the master-side storage device is installed, insofar as the damage does not reach the remote-side storage device located in the remote place, the system as a whole is able of operate continuously simply by effecting a changeover to the remote-side system.
Generally, since the remote copy to a remote place takes time in data transmission, the master-side storage device issues a response of reception complete upon storage of the data in an internal buffer, and subsequently effects data transmission to the remote-side storage device asynchronously with host transmission.
At present, the above-described technique is realized even in the midst of the input or output of the data, and the starting of the copy to the remote side or a changeover of the system is possible without stopping the online system.
In recent years, the transfer performance between host computers and master-side storage devices continues to be improved remarkably as exemplified by the fact that Fibre Channel Protocol has been standardized by ANSI and NCITS T11. On the other hand, channel resources for connection between master-side storage devices and remote-side storage devices are limited for economic reasons, and cannot be expanded unlimitedly. In particular, when a system having the remote copy function is structured, if a measure against wide-area disasters is taken into consideration, it is desirable to install the remote-side storage device farther away from the master-side storage device. However, if remote copy to a remote place is effected, even if a public line is used or a private line is laid, huge costs are incurred.
As a result, in comparison with data transfer channels between the host computer and master-side storage device, remote copy channels between the master-side storage device and the remote-side storage device are inevitably limited in number, and the marginal performance of data transmission is relatively low.
Accordingly, for example, in the event that the storage of data from master-side host computers occurs in large volumes in excess of the marginal performance of the remote copy channel, the remote copy cannot follow the speed, so that uncopied data overflow from the data buffer inside the master-side storage device. If the data buffer of the master-side storage device becomes full, acceptance of new writing of data from the hosts is prohibited to prevent the occurrence of the loss of data. However, it becomes impossible to store data from the other host computers just when a large volume of data flows from one of the plurality of host computers connected to the master-side storage device.
In the case where a plurality of host computers are connected to the master-side storage device, all the host channels do not necessarily have the same degree of importance. It is conceivable that there are variations in the performance requirements of the host channels and in the amount of data transferred. For example, trouble can conceivably occur that even when data storage occurs in a large volume from a host channel whose performance requirement is low or which is relatively unimportant, and the above-described phenomenon of the full data, buffer occurs, the writing of data from other host channels which are important becomes impossible, and this has been a major technical problem in operation.
Meanwhile, in recent years, in conjunction with the improvement of performance of personal computers, workstations, and the like with general-purpose operating system (OS) for small computers installed, these small computers have come to be connected to storage subsystems as host computers in addition to the mainframe.
For this reason, it is conceivable to adopt a form of operation in which, on the storage subsystem side, in addition to the mainframe-system host interface, the so-called open-system host interface for small computers is provided to allow both the mainframe-system and open-system host computers to share the storage subsystem.
In that case, it is not realistic to connect a plurality of storage subsystems for effecting remote copy of data by the above-described diverse interfaces due to the economic limitation of the remote copy channels as described above, and there is a technical problem in that it becomes necessary to uniformalize the data format among the different host connection interfaces.
SUMMARY OF THE INVENTION
Accordingly, it is an object of the present invention to provide a storage subsystem capable of effecting remote copy of write data among a plurality of storage subsystems with the order of priority corresponding to the degrees of importance of the plurality of host computers connected to the storage subsystems, as well as its controlling method.
Another object of the present invention is to provide a storage subsystem capable of realizing efficient remote copy among a plurality of storage subsystems without being affected by an increase in the load of data writing by a specific host computer among a plurality of host computers connected to the storage subsystems, as well as its controlling method.
Still another object of the present invention is to provide a technique capable of structuring an efficient remote copy system by uniformalizing diverse connection interfaces between a plurality of host computers and storage subsystems to a connection interface among the plurality of storage subsystems for effecting the remote copy of data.
In accordance with one aspect of the invention, there is provided a storage subsystem including a first storage subsystem connected to a plurality of host computers via a first interface and a second storage subsystem connected to the first storage subsystem via a second interface so as to copy write data written in the first storage subsystem from the host
Fujimoto Takeo
Kishiro Shigeru
Hitachi , Ltd.
McDermott & Will & Emery
Nguyen Hiep T.
LandOfFree
Storage subsystem and its controlling method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Storage subsystem and its controlling method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Storage subsystem and its controlling method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3090876