International Business Machines Corporation

Amal A Shaheen-Gouda

Todd A Smith

Donavon W Johnson

Grover Herbert Neuman

Larry K Loucks

Gerald G Barrett

Syed Z Pasha

Stephen P Morgan

Marion Lee Blount

Albert Y Chang

Larry W Henson

Katalin A Rader

Robert K Rader

Charles H Sauer

A unique naming system and method are described for managing object identification by a network of computer systems. The naming system employs data structures stored in the memory of the computer systems containing character strings and corresponding addresses to entries in the data structures and objects in the computer systems. Names employed in a particular computer system that correspond to objects in another computer system are resolved by a border data structure capable of transforming names, delimiters and visas across computer systems.

Name resolution for a multisystem network

A communication protocol for direct communications between operating system kernels of a plurality of processors in a clustered-type, shared virtual memory, multi-processor data processing system is disclosed. Each processor unit includes a intermachine communications software system ('IMCS') which may-be used by a trusted kernel of the operating system of the processor to communicate with a service in another processor, either directly or through an intermediate processor. Communications are initiated when a user prepares a message in accordance with pre-programmed instructions from the IMCS in a standard format. Once the message is prepared, the user calls IMCS which selects a route for the transmission of the message over a serial communications link to the receiving processor using a 'request to send immediate' serial link level protocol. The message is initially directed to a common buffer pool. IMCS calls special software to read the header of the message to determine the identity of the destined server and to determine if the buffer in the buffer pool assigned to the called service has available buffer space to accept the message. If not, IMCS directs the sending of a negative acknowledgment to the sending processor. If sufficient buffer space is available, IMCS directs the sending of an acknowledgment to the sending processor. Upon receipt of the acknowledgment by the sending processor, IMCS calls a notification routine to inform the user of the status of the transmitted message.

Intermachine communication services

A conventional single node operating system is provided with a distributed file management system (DFS) with a plurality of nodes and a plurality of files. The DFS uses the UNIX operating system tree structure employing inodes (data structure containing administrative information for each file) to manage the local files and surrogate inodes (s.sub.-- inode) to manage access to files existing on another node. In addition, the DFS uses a lock table to manage the lock status of files. The method which implements the DFS locking of records and files involves the following steps. If the file is a local file, then the UNIX operating system standard file locking is used. However, if a remote file is to be locked, the UNIX operating system LOCKF and FCNTL system calls are intercepted and an remote process call (RPC) DFS.sub.-- LOCK.sub.-- CONTROL is executed. The server node receives the remote process call and carries out the lock request. The request could entail locking a single record, a set of records or the whole file. The server then acknowledges receipt of the RPC by sending a signal while the client surrogate inode is waiting for a reply from the DFS.sub.-- LOCK.sub.-- CONTROL RPC. The client confirms the reception of the lock and sends an acknowledgement to the remote server. The server updates the lock table after receiving the acknowledgement from the client surrogate inode. If the server does not confirm the reception of DFS.sub.-- LOCK.sub.-- CONTROL's acknowledgement, then DFS.sub.-- LOCK.sub.-- CONTROL updates the lock table.

System for file and record locking between nodes in a distributed data

A distributed file management system (DFS) with a plurality of nodes and a plurality of files is disclosed. The DFS uses the UNIX operating system tree structure employing inodes (data structures containing the administrative information of each file) to manage the local files and surrogate inodes (s.sub.-- inode) to manage access to files existing on another node. In addition, the DFS uses a file access structure lock (fas.sub.-- lock) to manage multiple requests to a single file. The primary reason for the addition of the fas.sub.-- lock for each file is to avoid the problem of deadlocks. The inodes and s.sub.-- inodes use the fas.sub.-- lock to synchronize their accesses to a file and avoid a deadlock situation where both s.sub.-- inode and inode await the use of a file that is locked by the other.

Distributed file access structure lock

An improved directory caching technique is provided for a plurality of data processing systems which are connected together in a network. In the system, when a local, or client, data processing system interrogates a remote, or server, data processing system for a unit of directory information, the server system is enabled to automatically send additional units of pertinent director information back to the client system in response to a subsequent change in the directory structure of the server system. If the server system is unable to continue updating the client system, for any of a plurality of possible reasons, the server system informs the client system of this fact, which enables the client system to purge itself of the formerly stored directory cache entry relative to this path, since the client system can no longer consider this cached path information to be currently correct.

Directory cache management in a distributed data processing system

A data processing system including at least one storage device for storing and retrieving data and connected to a storage device interface that controls the storing and retrieving of data from the storage device and further includes a data buffer for storing data retrieved from or being stored in the storage device. The data processing system includes a storage device driver that receives storage access commands from a processor, provides commands to the interface in response to these commands to the access commands from the processor and also provides commands for data not requested by the processor. The commands for data not requested by the processor provide for the temporary storage of data in the storage device interface data buffer. This data not requested by the processor is retrieved in anticipation of commands to be sent by the processor. The storage driver manages this buffer to optimize the amount of storage dedicated to the storage of data for these anticipatory commands.

Memory disk buffer manager

A system for an efficient message handling technique implemented in AIX, an operating system derived from UNIX System V, is described for use in a distributed services network include a plurality of multi-processing, multi-tasking nodes among which interprocess communication occurs via queues, the actual node locations of which are transparent at the application interface.

Interprocess communication queue location transparency

A data processing system including at least one storage device for storing and retrieving data from several rotating tracks where each track includes sequentially located blocks for the storage of data. A storage interface is connected for controlling this storage device. The storage interface includes a data buffer for storing data retrieved from the storage device. A storage driver is provided that receives storage access commands from a processor and provides commands to the interface in response to these commands from the processor. This storage driver further includes the capability to provide commands for requesting data that has not been requested by the processor. This data not requested by the processor is stored in a data buffer. These commnads for data not requested by the processor are issued in accordance to procedure that computes these commands called read ahead commands based on the order of commands received from the processor and also based on the number of storage devices in use.

Memory disk accessing apparatus

In a distributed environment several data processing systems are interconnected across a network system. A distributed services program installed on the systems in the network allows the processors to access data files distributed across the various nodes of the network without regard to the location of the data file in the network. The processing system accessing file, referred to as the client processing system, utilizes a client cache within its operating system to store the data file. Utilizing the client cache minimizes the number of reads and writes that must go over the network to the server processing system where the file physically resides. The system and method of this invention prevents a process in the client processing system from accessing data in the client cache that has been modified at another node in the network. The blocks of data in the client cache are tested for validity in the client processing system by using modification times as measured by the server processing system. If the cache data blocks are determined to be valid, the data blocks are determined to be invalid, the data blocks are discarded, and the file is accessed from the server processing system.

System and method for using cached data at a local node after re-opening

A distrbuted services program installed on each of a plurality of data processing systems in a network allows the processors to access data files distrbuted across the various nodes of the network. To reduce the network traffic overhead when files at other nodes are accessed, and to preserve the file system semantics, i.e. the file integrity, the accessing of the various files are managed by file synchronization modes. A file is given a first synchronization mode if a file is open at only one node for either read or write access. A file is given a second synchronization mode if a file is opened for read only access at any node. A file is given a third synchronization mode if the file is open for read access in more than one node, and at least one node has the file open for write access. If a file is in either the first or second synchronization mode, the client node, which is the node accessing the file, uses a client cache within its operating system store the file. All read and writes are then sent to this cache. If a file is in the third mode, all read and write requests must go to the server node where the file resides. The node accessing the file does not use the cache in its operating system to access the file data during this third mode.

Growing community of inventors

Austin, TX, United States of America

Amal A Shaheen-Gouda