JPH0962524A - Method and system for controlling event in dispersed operating environment - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide event managing service(EMS) which is integrated without interruption while using an existent authentication, event transmission and remote procedure call(RPC) mechanism in the distributed arithmetic environment. SOLUTION: An event channel is provided to asynchronously apply events from one or plural event providing processes to one or plural event using processes by the EMS. The event is transmitted by generating an RPC. It is preferable to authenticate the event providing process and using process when registering them on the EMS.

Description

Detailed Description of the Invention

[0001]

FIELD OF THE INVENTION This invention relates generally to computing networks, and more particularly to distributed computing environment domains.
On a method for managing events in.

[0002]

BACKGROUND OF THE INVENTION It is well known that by interconnecting multiple computers as a local area network (LAN), these computers can exchange information and share resources. Local Area Networks provide a distributed computing environment,
Within this environment, users can access distributed resources and process applications on multiple computers. An "event" is an indication of a change in the state of a given managed resource in this environment.

A well-known distributed computing environment is called DCE and is realized using software available from the Open Systems Fundation (OSF). In a conventional OSF DCE computing network implementation, local
Provide event service only on level. Therefore, events that occur on a particular machine in the network are only interacted with and displayed on that machine itself. Other machines in the network do not provide an efficient means for receiving notifications when events have occurred on remote resources within the network. This drawback makes it difficult for system managers to remotely manage other resources in the network.

There are several event service architectures known in the art. The first is DME Event Service (DES), which is based on Ferrante.
By DME Event Service Architecture, Version 2.
0 ”(1993/4/28). This service architecture uses the concepts of an event providing process (a process that generates event data) and an event using process (a process that processes event data generated by a providing process), but has a complicated indexing mechanism. Due to its complexity, it cannot be incorporated into other existing DCE services. Another service architecture is the OMG event service model,
For more on this, see Joint Object Services Submission.
-Event Service Specification '' (OMG TC 93.7.3, 1993
/ 7/2). This service architecture provides two object-based event services and defines two basic models for communicating event data. One is the push model,
Allow the event-providing process to start forwarding events to the event-using process. The other is pull
A model that allows an event-using process to request an event from an event-providing process. Theoretically, event channel objects allow multiple providing processes to communicate with multiple consuming processes in an asynchronous manner. The main drawback of the OMG event service model is that there are no specific implementations currently available.

In both conventional and object-oriented architectures, there remains a need to provide event management services that allow managed resources to send asynchronous notifications or "events" to a group of interests. ing.

[0006]

The main object of the present invention is to:
Multiple event providing processes in a distributed computing environment
It is to be able to communicate with multiple event usage processes in an asynchronous manner.

[0007] Another object of the present invention is to closely integrate the event management service into an existing distributed computing environment,
It is possible to transparently supply an event in an asynchronous manner to a permitted remote event using process.

A further object of the present invention is the DCE remote
To deliver an event between an event providing process and an event using process by generating a procedure call (RPC).

A further object of the present invention is to provide a DCE / RPC-based asynchronous event management service in which the event providing process and the event using process are authenticated before using the service provided by the DCE security service.

A further object of the present invention is to integrate event management services into embodiments of DCE networks that use DCE network event transmission, security, and remote procedure call services.

A further object of the present invention is to filter events generated by one or more providing processes to determine if a particular event was communicated to one or more event consuming processes. Is.

It is a further object of the present invention that the event usage process is located and registered by one or more event management services running on multiple host computers in the network. It is possible. The event consuming process can control the host computer that sends the event and the particular type of event that is sent.

A further object of the present invention is to manage events so that the event providing process does not recognize an event using process registered as a receiver of the event. Preferably, a particular providing process only needs to send an event once in order to ensure that the intended event consuming process receives notification of the event.

A further object of the present invention is to reliably deliver events to the consuming process with minimal network and system load, and in a manner that can survive network and machine failures.

[0015]

In one example of a method for managing events in a distributed computing environment, one or more event providing processes generate events used by one or more event using processes. The method begins by an event consuming process registering with an event management service (EMS) to receive selected events originating from one or more event providing processes. In response to a Remote Procedure Call (RPC) from the DCE Event Recording Service, the Event Management Service uses an R
Determine whether an event associated with the PC has been selected by the event usage process. If the EMS determines that the event data associated with the RPC has been selected by the event consuming process, the EMS issues an RPC to supply the event data to the event consuming process. When multiple event-using processes are registered with EMS, the events occurring on remote machines in the network are asynchronous and are delivered transparently to the event-using processes.

In the preferred embodiment, the event management service is DC
It is tightly coupled to the E-event recording service and this integration is not necessary, but allows direct access to the event management service. further,
The providing and using processes (although not required)
It is preferable to be authenticated in cooperation with being registered in the event management service, and it is effective that such authentication is performed by the DCE security service.
Therefore, the event management service preferably utilizes the existing event recording, security and RPC services in the DCE network to provide efficient asynchronous event management.

Another unique feature of the present invention is that
A new filter mechanism is provided. this is,
It controls how the event data generated by the event providing process is passed to a particular consuming process. According to this aspect of the invention, each event consuming process registered with the event management service determines whether a particular event generated by one or more event providing processes is communicated to that event consuming process. Specify the event filter group to be executed. The event filter group includes one or more event filters, each of which is derived from information in one or more predetermined event type schemas and / or headers associated with each event. The filter mechanism parses the event through the event filter group of the event consuming process. Events that satisfy the parsing protocol for a particular event consuming process are communicated to that consuming process via RPC.

In event services, multiple event-using processes are often subject to the same event data, but event-using processes cannot use events concurrently with transmission from the event management service (EMS). When sending further events, while the providing process is trying to send the event to the EMS,
It should be unobstructed. And the EMS should be unobstructed while trying to send the event to the intended usage process. To achieve these goals, EMS implements a new queuing mechanism that utilizes a number of threshold-based services. Such services include process control routines that ensure that no event-providing process or any event-using process takes any action that interferes with the event channel.

The above is a summary of the objects related to the present invention. It is to be understood that these objectives merely indicate some of the more prominent features and applications of the present invention. Many other useful results can be achieved by applying the disclosed invention in various ways or by modifying the invention described below. Accordingly, other objects and a complete understanding of the present invention will be obtained from the detailed description of the preferred embodiments below.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a known distributed computing environment (D
CE), showing two or more nodes A, B and C connected via a communication link or network 3.
including. The network 3 is a local area network (LAN) or a wide area network (W
AN), the latter consisting of switched or leased line teleprocessing (TP) connections to other nodes or networks of the system operating under known computer architectures. Nodes A and B
Or C, the processing system 10A, 10
B or 10C is present. Each of these systems
It may be a single-user system or a multi-user system. Each processing system acts as a client or a server depending on whether it requests or provides a service.

Each processing system is a computer, and may be referred to as a "machine" in this specification. For example,
Each computer is a RISC System / 6000 ™ (short instruction set or so-called RISC-based workstation), and AIX ™ (Advanced Interactive Execu
tive) operating system. A
The IX operating system is AT &T's UNI
Compatible with the X operating system, version 5.2 at the application interface level. Various models of RISC-based personal computers are described in many publications from IBM, such as "RISC System / 6000, 7073 and 7016 POWERst.
ation and POWERserver Hardware Technical Referenc
e, Order No. SA23-2644-00 ”. The AIX operating system is "AIX Operating System Technic
al Reference "(published by IBM Corporation, 1st edition 1985/1
1) and other publications. For a detailed description of the UNIX operating system design, see Maurice J.
"Design of the Unix Operating System" by Bach
(Published by Prentice-Hall, 1986). As another example, each computer may be an IBM ™ PS / 2 ™ operating under the OS / 2 ™ operating system. For more information on PS / 2-based computers and OS / 2 operating systems, see the Technical Reference Manual Personal Systems /
2 Model 50,60 Systems IBM Corporation '' (Part No. 6
8x2224 OrderNumber S68X-2224) and `` OS / 2 2.0 Techni
cal Library, Programming Guide Volumes 1-3 Version
2.00 ”(Order Nos. 10G6261, 10G6495 and 10G6494).

In a particular embodiment, the present invention provides an IB
TCP / IP (Transmission Control Protocol / In)
It runs on multiple IBM RISC System / 6000 machines interconnected by the ternet Protocol) architecture. T
CP / IP uses Ethernet as its link level, which is a local area network (LAN) developed by Xerox. For a brief description of local area networks, see Larry E. Jordan and B.
`` Communica itons and Network in
g for the IBM PC '' (Prentice-Hall, Robert J. Brad
y Publication 1983). While the present invention has been described in the above context, the teachings herein are based on the
It should be understood that it could be implemented with other different computers interconnected by other networks than rnetLAN or IBM's TCP / IP.

Preferably, the present invention is implemented in a distributed computing environment having any number of machines, which machines may be located in different geographical locations. For purposes of explanation, the following detailed description is directed to the DCE domain or "cell". Although this is a relatively small network (up to about 25 machines), the present invention is applicable to any distributed computing environment (whether conventional (SNMP and CMIP) or object oriented). FIG. 2 is a diagram showing a DCE cell 12 embodying the present invention. The cell 12 consists of a set of connected machines and includes at least one server 14 and a DCE client 16, which share a common cell name and namespace. Each cell provides a location independent namespace service, referred to as the CDS or Cell Directory Service 15. Naming services are used by application servers to store their location and interfaces and are known as server bindings. A cell typically includes one or more other server machines that provide other services in this distributed computing environment. These other services include security services that support authentication 1.
8 and the application developer
It includes, but is not limited to, the Event Recording Service 20, which is a DCE service that can be used to record events.

In known DCE implementations, events have several types. "Service" events are usually
Includes error messages, warnings, abort messages, general resource information, etc. Another type of event is an "audit" event. Known DCE implementations include an SVC (Serviceability) interface and an audit interface within the event recording service 20 to handle these events. The SVC interface and audit interface can record events.

According to one aspect of the invention described below, calls to the SVC and audit interfaces within the event recording service 20 are made via the remote procedure call (RPC) to the event management service 22.
Automatically given to. This is also a server call “emsd”, and so-called “event providing process” and “event using process” can easily perform asynchronous management of events between them. EM
S22 can also be called directly. For convenience, event passing through the SVC or audit interfaces is often referred to as first level or "providing process side" event filtering. Generally, an event providing process is a DCE-based application (DCE core (co
re) -including the server). "Event" means a state change of a given managed resource and is embodied as data transmitted between the event providing process and the EMS and between the EMS and one or more event using processes. To be done. Such event data consists of an event header and a list of event attributes containing event format specific data. In one embodiment, all DCE applications that use the SVC interface or audit interface of the event recording service 20 are event providing processes. "Event usage process" means
Any DCE server application registered in EMS 22 (discussed below) that receives and processes events from one or more "event providing processes".

FIG. 3 is a block diagram showing a preferred example of the event management service of the present invention in an existing DCE system. DCE subsystems and functions are shown,
SVC interface 30 and audit interface 2
8 and Remote Procedure Call (RPC) 3
Including 1. To reach the event management service 22, the event providing process can use the EMS application programming interface (API) 32. The EMS 22 acts as an “event channel” and one or more event providing processes 24a.
24n to one or more event using processes 26a to
26n. Communication through the event channel is "asynchronous" in that it can be given to the event consuming process at any time. The event consuming process can be a local process as shown or a remote process. Communication is performed between the event providing process and the EMS on the one hand, and between the EMS and the event using process on the other hand using the RPC mechanism of the DCE system. The portion drawn within the dashed line is for a single host in a distributed computing environment (for illustration purposes only).
Reflects the computer, EMS22 and multiple DC
There is an E event providing process 24. The event recording service 20 includes an audit trail 34, S
Including a VC log file 36 and a console 38,
The management function is executed here.

As will be described in more detail below, the event management service includes a number of databases and memory structures. That is, the used process database 40, the event log file 42, the event format database 44, and the event filter database 46.
It is. In order to start receiving events, the event usage process must first be registered with the EMS 22. One of the typical techniques for registering an event use process is
After giving a handle that describes the connection to the service,
The event consuming process uses that handle to generate an RPC to the EMS. The EMS then places an entry in the usage process database 40 that uniquely identifies the usage process. "Event provision process"
Is registered with the event management service by receiving the handle. The registration process thus defines a connection to a particular EMS process running on a particular machine. Although not required, the event-using process and the event-providing process in registration may authenticate themselves using a DCE security service (not shown).

The event usage process registered in the EMS is identified in the usage process database 40.
Event log file 42 stores a collection of events to be sent to the event consuming process. The event log stores an event when the EMS cannot send the event to all appropriate event consuming processes (eg, due to a network connection being lost). The event is deleted from the event log after it has been sent to all relevant usage processes.

However, the event consuming process requires that a "filter" be defined before event data can be received. Then use this to EM
S determines whether a specific event from one or more event providing processes is to be passed to the event using process. First, the event consuming process uses general "event type" information in the event type database 44 to create a particular "event filter group" for the event consuming process. As shown, the information in the event type database 44 is also used by the event provision process and possibly the EMS administrator. The filter data selected or identified by the event usage process is then stored in the event filter database 46.

The event format database 44 is EMS.
It also stores the information used by the event providing process and the using process and generates an "event format".
Each event type is a class of events that have the same event type format. The event format is described via an "event format schema" and is identified by a unique universal identifier (UUID). The event format schema is a description of the event format, and consists of a list of pairs of attribute names and attribute formats that specify the data format of the event. An attribute name is a string that uniquely identifies an attribute of a given event type. The attribute format is a data format of the event attribute and defines the format of the data of the attribute. The event usage process uses the event format schema data (stored in database 44) and / or event header information to create an “event filter group”.
To build. The "Event Filter Group" determines which events should be sent to its consuming process.
Tell S.

FIG. 4 shows that an event filter group is a collection of one or more "event filters" that are logically OR'ed together. There is one event filter group for each registered event usage process. "Event
A "filter" is a collection of one or a plurality of "filter notations", and is logically connected to each other by a logical product (AND). And each event filter has an event filter name and event type (which is common)
Have. Any desired number of event filter names can be added to the consuming process's event filter group, which controls which events the EMS sends to that consuming process. The event filter name is listed in the using process database. The "filter notation" of an event is preferably a set with three elements: an attribute name, an attribute value, and an attribute operator that defines the comparison operation. The filter notation attribute operator is used to act on the comparison between the named attribute and the attribute value of the event. The attribute name in the filter notation is
It may mean the attribute name of the event format schema or event header information. Thus, for example, if the header information includes time stamp information, the event filter group may include filters with event filter notation meaning time stamp information in the header.

Referring again to FIG. 3, in order to reach the EMS, an event is sent via the event recording service to the EMS.
Is passed to or EMS is called directly. When an event reaches the EMS via RPC, it is stored in the event log 42. The EMS 22 then determines whether to pass the event to any event consuming process by performing a parsing operation. The information stored in the event filter database 46 is the second stage or "using process" EM.
The S filter 45 is used by the EMS 22 to perform this evaluation. The second stage filtering is also shown in FIG. After the filter processing, a queue mechanism 47 that controls the flow of events to the target usage process is used (described later). After sending the event to all the target usage processes, the event is recorded in the event log 42.
(And queue).

In the filter processing routine, the event is EM.
It is started every time S is reached. At step 52 of FIG. 5, the EMS retrieves the event format schema for the particular event from the event format database. In step 54, the EMS 22 lists the registered event usage processes in the usage process database 40. For each registered usage process,
The following processing occurs. In step 55, first, the event filter group of the registered usage process is retrieved from the usage process database. As mentioned above, each event filter group means a collection of one or more filters, each containing a list of one or more filter notations. After that, the parsing protocol is started. In particular, in step 56 it is determined whether all the filters in the event filter group have been evaluated. If not, go to step 58 to evaluate the next filter in the event filter group. Examine each filter notation within each filter. Then, in step 60, a test is performed to determine if all filter expressions within a particular filter are in a preset state or "TRUE". If all filter expressions are true, then
Passes the event to the RPC mechanism for delivery to the registered event consuming process. Step 60
If any of the filter expressions in the filter is “FALSE” as a result of the inspection in step 56, step 56
Return to get the next filter in the event filter group (identified by the next event filter name). If the check at step 56 indicates that all filters in the event filter group for this consuming process have been evaluated, then this routine is exited and the EMS proceeds to the next consuming process. In this way, according to the parsing protocol, the EMS will not send an event to the event consuming process unless all filter expressions of a particular filter in the event filter group are true.

Still referring to FIG. 3, the consuming process API may use the consuming process API to define a new event filter and add it to the event filter group. The consuming process may also take a list of existing event filter names and add one of them to the event filter group. The consuming process can add event filter names (and their filters) to and from the event filter group (through the consuming process API). Further, the EMS management mechanism can be similarly operated through the management API shown in the figure. In this way, the event consuming process can modify the event filter group by adding or deleting new filters.

The event format schema stored in the event format database is preferably used in several different ways according to the invention. The consuming process may request a list of supported event types and select which event types it wishes to receive using the event type schema to create an event filter. The event consuming process can further map events according to attribute names. The event providing process defines a new event format to be generated by using the event format schema. EMS uses the event format schema to apply event filters to events.

In many cases, even though multiple event-using processes are targeted for the same event, these event-using processes may not be able to use the event immediately when the event is transmitted from the EMS. is there. In addition, when sending an event, the providing process should not be disturbed while trying to send an event to the EMS, and should not be blocked while trying to send an event to the consuming process that the EMS is intended for. To achieve these goals, the EMS implements a queue mechanism 47. The queue mechanism 47 has a queue that defines a data structure and a set of queue operation routines. These routines preferably use another existing DCE service, the DCE thread, to
E Enables event management services to manage event queues. As is known in the art, threading allows multiple routines to execute conceptually simultaneously and share resources. DCE Thread Services is a set of routines that can be used to create multithreaded programs. Multithreading is used to improve the performance of a program by allowing pieces of the program to operate simultaneously. The DCE thread is "Draft 4 Of The POSIX1003.4
It is a user-level (non-kernel) thread package based on the PTHREADS interface defined in "Standard".

This event queue mechanism is now generally described. According to the present invention, in FIG. 6, as long as the event using process is registered in the EMS 22, the using process queue 70a for the using process is set in the EMS. As shown in FIG. 6, the consuming process also has a corresponding event queue 7
2a. The event service is the input queue 74
And an activity queue 76. According to the present invention, each queue has a number of queue elements, and each queue element can have either of two data formats. One is a pointer to the event and the other is a pointer to another queue element. EM
The activity queue 76 of S has elements that indicate events that have not been sent to all of the intended use processes (ie, the use processes registered to receive the current event), each queue element being Corresponds to one event. In addition, each queue element in the active queue has a mutex protection count that indicates the number of associated consuming processes that have not yet received the event. The EMS use process queue 70 has elements that contain pointers to the elements in the activity queue 76. The appropriate queue mechanism routines
Create and initialize a queue, destroy the queue, enqueue (or "push") elements to the queue, dequeue (or "pop") elements from the queue, remove elements from the queue, and delete the queue (After recovery)
Activate.

FIG. 7 is a diagram showing in detail the queue management routine. In step 80, the routine begins by setting up the consuming process queue 70 for each registered event consuming process. When sending an event to EMS 22, the event is immediately placed in input queue 74 and the RPC routine returns. This is step 82 in FIG. (Input queue
The queue mechanism routines (referred to as process elements) are running in a separate DCE thread, but then signaled at step 84. This indicates that the input queue 74 is not empty. At step 86, each element is dequeued from the input queue 74. If any consuming process is the subject of the element's directed event, then at step 88 a pointer to the element is enqueued on the consuming process's consuming process queue 70 (in EMS), and The count associated with the element is incremented (indicating that another event consuming process is registered to receive the event). Step 9 (after the used process queue 70 of the target used process is updated by receiving the pointer)
At 0, the queue element for the event is enqueued on the active queue and waits until all consuming processes have received the event. At this point, the count represents the number of associated event consuming processes (ie, consuming processes seeking notification) that the EMS must notify of the event.

When an element is enqueued in one of the EMS's use process queue 70, the thread executing the use process queue element routine in step 92 receives a signal. Then, in step 94, the routine dequeues the element from the consuming process queue and sends it to the associated consuming process queue. Once the queue element has been sent, the count associated with that element (in the active queue) is decremented in step 96. If the count reaches zero, then all target usage processes have been posted and the queue element is removed from the active queue at step 98.

Each of steps 92, 94, and 96 comprises:
Executed in separate threads, multiple instances of these steps occur simultaneously (at least conceptually) with respect to the consuming process queue. When all threads have successfully sent a particular event, that event is removed from the active queue (and event log 42).

As mentioned above, each consuming process also has an event queue 72, as shown in FIG. When an event is sent to the consuming process, the event is put into this queue and a signal is sent to the process element routine. This routine is now Event Queue 7
Dequeue the element from 2 and call the event handler to retrieve the event.

Next, a more detailed specification of the event management service (EMS) implemented in the OSF DCE will be presented. As mentioned above, DCE EMS manages event services in DCE cells. Preferably EM
S is an ems to access the event service
It consists of two main parts: d-server and API. em
sd is a DCE server that can exist on any DCE host in the cell, from which the using process requests events. It is also envisioned that emsd is implemented in a distributed fashion. Three interfaces have been exported by emsd to support the EMS provisioning process, consuming process, and event service management mechanism. The EMS API provides interfaces to the same three interfaces used by EMS clients.

Hereinafter, the operation mode of EMS, the description of the definition method of the event format, the flow of the event via EMS, the operation mode of the event filter, the event logging method, and both the event data and the event filter definition will be described. The security model used to protect is described in detail.

<Explanation of Event Flow> As described above,
The EMS sets up an event channel to separate the communication between the providing process and the consuming process. To start receiving events, it is preferred to first register the event usage process with the EMS. Then an event filter, to tell EMS which events should be sent to its consuming process.
Set up a group. Before the providing process can send an event to the EMS, the event is usually passed through a first stage event filter. In this example, the first stage filtering is implemented at the level of the providing process by using the SVC mechanism and the audit mechanism. Once the event has passed the first stage filter, it is sent to the EMS. EMS is
If the event cannot be delivered immediately, write it to the EMS event log to save the event.

When an event reaches the EMS, it is passed through a second stage filter before being sent to the intended usage process. Preferably, the EMS, after listing the list of registered usage processes, retrieves the event format schema from the event format database, the event filter group of the usage process from the usage process database, and the event filter database. An event filter is used to determine whether this event should be passed to be sent to each event consuming process. The event is removed from the event log after all appropriate consuming processes have received the event.

<Event Format Definition> The EMS event format format is preferably defined by the event format schema and is held in the EMS event format database. An event format schema typically consists of a list of attribute names and an attribute format that specifies the data format of the data associated with that attribute. Each event is associated with a fixed header portion and a variable length data portion. The variable length data portion preferably consists of N self-defining data items, which is the attribute format or data itself. The data elements of the variable length part correspond in order, that is, they correspond one-to-one with the list of pairs of attribute names and attribute formats defined in the event format schema. The variable length portion can have further unnamed data items (ie, data items with attribute names) after named data items.

General-purpose event format EMS supports events having a general-purpose (Generic) format. Generic events do not have an event format schema. The only way to specify filters for generic events is to use filter notation with event header attributes.

Default event format The definition of the event attribute list included in the event format schema of SVC and audit is as follows.

[0049]

[Equation 1]

Several constants used to match against the severity (atttribute.severity) attribute are also defined. It is as follows.・ SVC_C_SEV_FATAL ・ SVC_C_SEV_ERROR ・ SVC_C_SEV_WARNING ・ SVC_C_SEV_NOTICE ・ SVC_C_SEV_NOTICE_VERBOSE

[0051]

[Equation 2]

<Event Filter> EMS supports the concept of two-stage filtering. The first stage filtering is applied to the machine executing the providing process and is a filter that is passed before the event is sent to the EMS. The second stage filtering is
EMS is applied to the event received from the providing process. This second stage controls which events are sent to the consuming process.

The following describes the two-stage filtering process supported by EMS, as well as how to define and configure an event filter.

Stage One and Stage Two Filtering In one implementation, SVC event formats and audit event formats are supported. The first stage filter for SVC events is the SVC routing destination. EM
SVC messages routed to S are sent through the EMS event channel. For auditing, audit records passing through the audit filter mechanism are sent through the EMS event channel.

The second stage filter is defined for each EMS usage process. This stage is
Associating a group with each event consuming process, and only that event consuming process receives an event that has passed one of the event filter group's entries for that event consuming process.

Defining Event Filters The normal situation where an event consuming process initiates an interface with an EMS is that the event consuming process determines which event format is supported by the EMS.
It is the first time to ask. From the list of event types, the event consuming process creates an event filter for that event type by querying the event type schema. The event format schema preferably includes a list of attribute names and attribute formats,
Using this, the event consuming process creates an event filter. The usage process then informs the EMS to apply the created event filter to the incoming event.

EMS filters can be created at several different levels. The lowest level is filter notation, usually attribute name, attribute operator,
It is a set of three elements consisting of attribute values that define comparison operations. All comparison operations are boolean. Only a given attribute operator is allowed a given attribute type.

Some examples of filter notation are shown below.

(Equation 3)

The first example evaluates to TRUE if the SVC event format attribute name "file" is the same as the string "file.c". In the second example, the attribute name "outcome" in the audit event format has the attribute value "aud.
The filter notation "xmp_Audit" evaluates to true if it is the same as "_c_esl_cond_denial". This means that the audit event records that access has been denied due to some action.

Event filters usually consist of a list of filter notations added together. Event filters have both a name and an event type. The event filter name is used to refer to the event filter during the create, delete, and update operations of the event filter. The event filter name is also used by the using process and the EMS management mechanism to add the event filter to the event filter group.

The event filter group of the consuming process is preferably one or more event filter groups.
It is a collection of names. As an event passes through the event filter group of the event consuming process, the event filter associated with each event filter name specified in that event filter group is applied to that event. If all event filter expressions in the named event filter evaluate to true, the EMS sends the event to its consuming process. If any of the event filter expressions evaluates to false, then the event filter specified by the next event filter name in that event filter group is applied to that event. Any event in that event filter group
If the filter also does not evaluate to true, the event is not sent to the consuming process.

Filtering of header information, that is, event format: Attribute name of a set of event headers set in advance /
Attribute format pairs are defined to allow filtering on the information in the event header. A filter notation with event header attributes can be a filter part with event format specific header attributes. However, a filter with a generic format can only have a filter notation with a header attribute.

<EMS Event Log> The EMS event log is used to store events in case of EMS failure. EMS will event all events
Write to the log, and delete the event record once the event has been transmitted to all consuming processes supported to retrieve it.

The event log is kept in a file on the machine on which emsd is running. Events are stored in the directory specified by the environment variable EMS_EVENTLOG_DIR. An API is provided to inspect the local event log.

<EMS Security Model> EMS
Secures the safe manipulation of the data in the EMS database. This includes a list of consuming processes in the consuming process database along with an event filter database, an event type database. EM
S further ensures the secure transmission of event data along with authentication and authorization of the providing and using processes. Every access control list (ACL) is associated with a name in the DCE namespace, and the emsd server manages the namespace following the next junction. /.:/hosts/<hosutname>/ems-server/ The ACL associated with this object controls access to emsd servers registered in this namespace. The permissions associated with the "ems-server" are:

[0066]

[Table 1] emsd server permission bit ------------- --- --- --- --- --- --- permission bit name description --- −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− c Change ACL to control server r Read Read attributes of this server s Stop EMS server Stop w Write Change the attributes of this server -----------------------.

The three security objects are
Maintained under the "ems-server" junction. The directories and databases they represent are as follows: event-types: Event format database filters: Filter database consumers: Usage process Each of these databases has an ACL associated with it.
Having.

Event Type Security Management The event type database is represented by the following names in the DCE namespace. /.:/hosts/<hostname>/ems-server/event-types The ACL associated with this object controls access to this database. The permissions associated with "event-types" are:

[0069]

[Table 2] Permit bits of event format database ----------------------------------------------- Permit bits Name Description --- −−−−−−−−−−−−−−−−−−−−−−−−−−−−−− c Change ACL for control server d Delete event type Delete schema i Insert event type Add a schema r Read Read the contents of the event format schema ----------------------------------------------------

Access to the EMS event data is permitted for each event format. Permission for event data for a given event type is /.:/hosts/<hostname>/ems-server/event-types/<event
It can be permitted by changing the ACL for _type_name>. Here, "<event_type_name>" is an event format name and appears in the event format schema. The following names are known as SVC and audit events. /.:/hosts/<hostname>/ems-server/event/SVC /.:/hosts/<hostname>/ems-server/event/Audit The permissions associated with "<event_type_name>" are:

[0071]

[Table 3] Permit Bits in Event Format ----------------------------------------------------------------------------------------------------------- Permit Bit Name Description ----- −−−−−−−−−−−−−−−−−−−−−−−−−−−−−− c Change ACL for control event type d Delete Delete this event type r Read this type Read (use) event of w write write (provide) event of this format ------------------------------------------- -----

The access rights of the providing process can be verified during the first event sent to the EMS, and the access rights of the consuming process can be verified before sending the event to the consuming process. Certified R
The PC is a remote A for the EMS providing process and the using process.
Used to access PI.

Event Filter Security Management The filter database is represented by the following names in the DCE namespace: /.:/hosts/<hostname>/ems-server/filters The ACL associated with this object controls access to this database. The permissions associated with the filters are as follows:

[0074]

[Table 4] Permit bits of the filter database -------------------- -------------------- Permit Bit Name Description ----- −−−−−−−−−−−−−−−−−−−−−−−−−−−−−− c Change ACL for control filter d Delete event filter Delete i Insert event Add a filter r Get a list or contents of the read event filters ---------------------------------------------------

Access control of the event filter is permitted for each event filter. The permissions for filter access for a given event filter are:
Allowed by changing the ACL of the next name. /.:/hosts/<hostname>/ems-server/filters/<filter_na
me> where "<filter_name>" is the event filter name given to the event filter during the call to "ems_filter_add". The permissions associated with the event filter are:

[0076]

[Table 5] Permit Bits of Event Filter ----------------------------------------------- Permit Bit Name Description --- −−−−−−−−−−−−−−−−−−−−−−−−−−−−−− c Change ACL for control filter d Delete event filter Delete w Write event Change the contents of the filter ------------- When the using process creates an event filter, The consuming process itself automatically gets the dwc permission for the created event filter.

-Use Process Security Management The Use Process Database is represented by the following names in the DCE namespace: /.:/hosts/<hostname>/ems-server/consumers The ACL associated with this object controls access to this database. The permissions associated with the usage process are:

[0078]

[Table 6] Permission bits of the used process database ------------------ Permission bits Name Description --- −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− c Control Change ACL for used process d Delete Delete used process i Insert Use Add (register) a process r Read List used process information w Write Change used process including filter group of used process. −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−

Initialization of EMS Security When the EMS is configured, some security groups are created by default. These groups are ems-admin (EMS management), ems
-consumer (EMS usage process) and ems-supplier
(EMS providing process). The default permissions are as follows. /.:/hosts/<hostname>/ems-server object ACL ems-admin crws hosts / <hostname> / self rws any_other r /.:/hosts/<hostname>/ems-server/event-types object ACL ems- admin cri ems-consumer r ems-supplier ri any_other r Initial object ACL (/.:/hosts/<hostname>/ems-s
erver / event-types / <event_type_name>) ems-admin cdw ems-consumer r ems-supplier w /.:/hosts/<hostname>/ems-server/filters object ACL ems-admin crdi ems-consumer ir any_other r initial Object ACL (/.:/hosts/<hostname>/ems-s
erver / filters / <filter_name>) ems-admin cw /.:/hosts/<hostname>/ems-server/consumers object ACL ems-admin cdrw ems-consumer irw any_other r

A new event created by setting these permissions for the ems-admin group
Each filter and event type will automatically inherit the same permissions.

The management mechanism adds all of the ems in the cell by adding itself to each of these groups.
It becomes possible to access the execution of d. If you want even stronger security, you can add yourself out of the group from each ACL.

When transmitting an audit type event, EM
S is an authenticated RPC set to DCE security protection level strong integrity (rpc_c_protect_level_pkt_in
tegrity) is preferably used.

Writing Usage Processes Events It is preferred that the consuming process be implemented as a server rather than a simple client. However, this is not essential. Each EMS usage process often has given requirements. For example, call consumer_start (start process in use), register event handlers, register with various emsds that wish to receive events, set up event filters, and rpc_server_lis.
It's ten's call.

A call to consumer_start creates a uuid for that consuming process and registers the EMS transmission with the consuming process handler. ems_consumer_handler_regis
The call to ter sets up the incoming process queue, initializes the event handler function and calls the specified handle function. One event consuming process can register to receive events from multiple emsd servers on different hosts. The consuming process must set up the event filter before the event is sent. Finally, the consuming process uses rpc_server_l to wait for incoming events.
Call isten.

<Recovery of EMS after Resuming> The EMS stores all the state data in the next repository. -Used process database Stores a list of all provided processes, and a list of filters in the used process filter group along with a list of all used processes. -Event format database Stores all known event format schemas. Event filter database Stores all current filter definitions. The following steps are performed when emsd is restarted. 1. All databases are restored. 2. All usage processes are notified of re-registration or re-connection. 3. The event queue is restored from the event log file. 4. Event transmission proceeds.

<Event Queue> The event queue has the following features. Empty and full variables: These variables wait until the queue is not empty or full, respectively. These variables can be turned on and off as needed. Queue mutex: used to enqueue and data queue elements on the queue to allow multithreaded access to the queue. Process Element Routine: This is a user-written routine with a defined code that is called to process the queue element. Maximum queue size: When the queue reaches this size, wait until the full state variable is no longer full.

Queue persistence: Each element on the queue is stored using the DCE backing store,
The event queue can be recovered if the EMS fails. Event queue element: The event queue element contains two types of data. A pointer to the event. A pointer to another queue element. This causes the EMS to have an activity queue with an element that indicates that the event has not yet been sent to all its target consuming processes, and for elements in the activity queue. A use process queue with an element containing a pointer. Each element also contains a mutex protection count that indicates how many pointers are active for that element.

<Data Structure> -EMS Event Attribute-Event Attribute Format The event attribute format is used to specify the data format of the event attribute. Attribute format is event attribute value combination
In (ems_attr_value_t), specify the format of the data. All event attribute formats are defined as follows. typedef unsigned16 ems_attr_type_t The event attribute format can be one of the following:

[0089]

[Table 7] Event Attribute Format Specifiers -------------------------------------------------------- Attribute Format Data Format Size- −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−− ems_c_attr_small_int idl_small_int 1 byte ems_c_attr_short_int idl_short_int 2 bytes ems_c_attr_long_int 8 bytes eml_per_int_r_per__r_per__r_per__r_hy_per____ 1 byte ems_c_attr_ushort_int idl_usmall_int 2 bytes ems_c_attr_ulong_int idl_ulong_int 4 bytes ems_c_attr_uhyper_int idl_uhyper_int 8 bytes ems_c_attr_short_float idl_short_float 4 bytes ems_c_attr_long_float idl_long_float 8 bytes ems_c_attr_boolean idl_boolean 1 byte ems_c_attr_uuid uuid_t 16 bytes ems_c_attr_utc utc_t 16 bytes ems_c_attr_severity_t ems_severity_t 4 bytes ems_c_attr_acl sec_acl_t ^* variable length ems_c_attr_byte _string idl_byte ^* Variable length ems_c_attr_char_string idl_char ^* Variable length ems_c_attr_bytes Structure reference Variable length −−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−−

Byte strings and character strings end with a 0 (zero) byte. The IDl compiler's pickling service can be used to encode complex data formats into byte strings that are included in EMS events.

Event attribute value The event attribute value combination is a self-defining data structure, and holds an attribute format specifier (format) indicating which format data is in the combination and the value of the specified data. With a suitable binding member. typedef struct ems_byte_s_t {unsigned32 size; / * size of byte data * / [size_is (size)] byte ^* data; / * byte data * /} ems_bytes_t; typedef union switch (ems_attr_type_t format) {case ems_c_attr_small_int; small int small_int; case ems_c_attr_short_int; small int short_int; case ems_c_attr_long_int; long int long_int; case ems_c_attr_hyper_int; hyper int hyper_int; case ems_c_attr_usmall_int; unsigned small int usmall_int; case ems_c_attr_ushort_int; unsigned short int ushort_int; case ems_c_attr_ulong_int; unsigned long int ulong_int; case ems_c_attr_uhyper_int; unsigned hyper int uhyper_int; case ems_c_attr_short_float; float short_float; case ems_c_attr_long_float; double long_float; case ems_c_attr_boolean; boolean bool; case ems_c_attr_uuid; uuid_t uuid; case ems_c_attr_utc; utc_t utc; case ems_c_attr_acl; sec_acl_t * acl; case ems_c_attr_byte_string; [string] byte * byte_string; case ems_c_attr_char_string; [string] char ^* char_string; case ems_c_attr_bytes; em s_bytes_t ^* bytes; default:;} ems_attr_value_t;

Event Attributes Event attributes typically include event attribute name / attribute type pairs, which define event attributes. Event attributes can be used in defining the event format in the event format schema and in defining the event filter in the event filter notation.
"Attr_name" specifies the attribute name, and "attr_type" specifies the attribute value format. typedef struct ems_attribute_s_t {ems_string name; / * Event attribute name * / ems_attr_value_t value; / * Event attribute format * /} ems_attribute_t;

EMS Event Structure Event Format The event format is the I-specific for a given event format.
Specify D. typedef uuid_t ems_event_type_t;

[0094]

[Table 8] Default event format -------------------- Event format Event format name --- −−−−−−−−−−−−−−−−−−−−−−−−−−−−− ems_c_generic_type General-purpose format ems_c_svc_type SVC ems_c_aud_type Audit −−−−−−−−−−−−−−− The general-purpose events have nothing to do with the event-format schema and can only be filtered by the header attribute notation in them. is there.

Event Identifier The event identifier uniquely identifies a given event.
Each event typically has a unique event type for all events of this type and a unique event ID for a particular event. typedef struct ems_eventid_s_t {ems_event_type_t type; / * event type * / uuid_t id; / * unique event identifier * /} ems_eventid_t;

Network Name The network name identifies the network name of a given host machine. Name services specify which name service recognizes a given network name. typedef enum ems_nameservice_s_t {ems_ns_other = 0, / * name service other than list * / ems_ns_dns, / * DNS name service * / ems_ns_dce, / * DCE CDS name service * / ems_ns_x500, / * X500 * / ems_ns_nis, / * NIS * / ems_ns_sna / * SNA network * /} ems_nameservice_t;

The structure of "ems_netaddr_t" specifies the actual network name. It can be interpreted according to the specified name service. typedef char ems_octat_t; / * used for NLS support * / typedef struct ems_netaddr_s_t {unsigned short len; [size_is (len)] ems_octet_t name []; / * name in proper format * /} ems_netaddr_t; typedef struct ems_netname_s_t_t {em service; / * name service used for net names * / [ptr] ems_netaddr_t * netaddr; / * network name / address * /} ems_netname_t;

For DCE host names, the following example sets the "ems_netname_t" structure. static char * dce_hostname = "/.:/ hosuts / eagle.austin.ibm.com"; ems_netname_t netname; netname.service = ems_ns_dce; netname.netadddr-> len = strlen (dce_hostname) +1; netname.netaddr-> name = (char *) malloc (netname.netaddr->len); strcpy (netname.netaddr-> name, dce_hostname);

Event Source The event source specifies from which event the event has occurred (that is, the providing process). The event source is the net name of the host on which the provisioning process is running, the name of the provisioning process or the descriptive name (descn
ame) and the provided process identifier (pid, uid, gid) are specified. These values may not be valid for all hosts. typedef struc ems_origin_s_t {ems_netname_t netname; / * Network name of the originating host * / [string] char * descname; / * Descriptive name of the providing process * / unsigned32 pid; / * Originating process ID * / unsigned32 uid; / * Source user ID * / unsigned32 gid; / * Source group ID * /} ems_origin_t;

Event Severity Event severity specifies the importance of the event. This name is one-to-one for DCE SVC severity attribute values
Corresponding to. typedef enum ems_severity_e_t {ems_sev_info = 0, / * Information event * / ems_sev_fatal, / * Critical event * / ems_sev_error, / * Caution event * / ems_sev_warning, / * Warning event * / ems_sev_notice, / * Notification event * / ems_sose_notice / * Verbose event * / ems_sev_debug, / * debug event * /} ems_severity_t

Event header The event header describes the fixed part of the event data structure. The event header contains the event ID, event originator, time and severity at which the event was received by the EMS and delivered to the using process typedef struc ems_hdr_s_t {ems_eventid_t eventid; / * event identifier * / ems_origin_t origin; / * event originator * / ems_severity_t severity; / * Event severity * / utc_t received; / * Event received timestamp * / utc_t delivered; / * Event delivery timestamp * /} ems_hdr_t;

The set of filter attributes is the event header
Provided for filtering. The following names can be used for filter attributes in the event filter notation.

[0103]

[Table 9] Event Header Attribute ------------------------------ Attribute Name Attribute Format -------- −−−−−−−−−−−−−−−−−− event.id ems_c_attr_uuid event.type ems_c_attr_uuid origin.netname.service ems_c_attr_ulong origin.netname.netaddr emc_c_attr_bytes origin.descname ems_c_attr_char_ul___________ origin.gid ems_c_attr_ulong severity ems_c_attr_ulong received ems_c_attr_utc −−−−−−−−−−−−−−−−−−−−−−−−−−−−−

The structure of the event “ems_event_t” is an event that is a fixed part.
It includes a header and an event data item that is a variable part. Each data item is a self-defined value that includes the attribute type and attribute date. The count identifies the number of data items that are in the event. typedef struct ems_event_s_t {ems_hdr_t header; / * fixed event header * / unsigned32 count; / * number of data items * / [size_is (count)] ems_attribute_t item []; / * data item * /} ems_event_t;

EMS Event Format The EMS event format structure is used to define the EMS event format.

Event format schema The event format schema is used to define the event format. It consists of an event format ID or format, a name field that identifies the name of the event format, and a list of event format attributes that describe the format of this event format. The size specifies the number of attributes in the event format. Only the event format schema identifies the fixed parts of the event. An event can have the same number of unnamed attributes according to the list of attributes specified here. typedef struct ems_event_type_scheme_s_t {ems_event_type_t type; / * EMS event format * / [string] char * name; / * event format name * / long size; / * number of attributes * / [size_is (size)] ems_attribute_t attribute []; / * Event type attribute * /} ems_evet_type_scheme_t

Event Format List The event format list contains a list of sizes of the event format schema. typedef [ptr] ems_event_schema_t * ems_schema_ptr_t; typedef struct ems_event_type_list_s_t {long size; / * number of event format schemas * / [size_is (size)] ems_schema_ptr_t schema []; / * ptr * /} ems_event_type_list_type_list_type_list_type_list_type_list_type_list_type_list_type_list

EMS Event Filter The event filter data structure allows the definition of both event filters and event filter lists.

[0109]

[Table 10] Attribute operator ------------------------------------------------------- Attribute operator Explanation of the attribute operator ----- −−−−−−−−−−−−−−−−−−−−−−−−−−−−−− ems_c_attr_op_eq true if attr_name is equal to attr_value (==) ems_c_attr_op_gt attr_name is greater than attr_value ( >) True true if ems_c_attr_op_lt attr_name is less than attr_value (<) true ems_c_attr_op_ge attr_name is greater than or equal to attr_value (> =) true True if attr_name is not equal to attr_value (<>) ems_c_attr_op_bitand True if the bitwise operation of attr_name and attr_value are greater than 0 true if ems_c_attr_op_substr attr_name contains the string value specified by attr_value − ----- −−−−−−−−−−−−−−−−−−−−−−−−−−−−− Attribute operators perform Boolean operations on attribute names and attribute values in event filter notation. Stipulate.
The attribute operator format is defined as follows. typedef unsigned16 ems_attr_op_t;

Event Filter Notation The structure of the event filter notation contains the elements of the event filter notation used to create the event filter. Event filter notation is attribute name,
A set of three values consisting of an attribute operator and an attribute value (att
r_name, attr_operator, attr_value)
Specify filter notation. typedef struct ems_filter_exp_s_t {[string] char * attr_name; / * attribute name * / ems_attr_ope_t attr_operator; / * attribute operator * / ems_attr_value_t attr_value; / * attribute value * /} ems_filter_exp_t;

Event filter notation list The event filter notation list groups the filter notations together into one list in order to form a filter notation by logical product (AND) when defining the event filter. . typedef struct ems_filter_exp_list_s_t {long size; / * number of filter notations * / [size_is (size)] ems_filter_exp_t filter_exps []; / * filter notations in list * /} ems_filter_exp_list_t;

Event Filter An event filter is a sequence of events that are ANDed with each other to perform a filter operation.
Identify the filter notation. Event filter is CD
The name that will be input to the S namespace (filter_na
me) and a list of filter expressions (event_exp_list). typedef struct ems_filter_s_t {ems_string_t filter_name; / * Event filter name * / ems_event_type_t type, / * Event filter format * / ems_filter_exp_list_t filter_exp_list; / * List of filter expressions * /} ems_filter_t; Filter with a generic event format Can only have a filter notation with the header attribute name in it.

Event Filter Name List The event filter name list contains a list of sizes of event filter names. typedef [string] char * ems_string_t; typedef struct ems_filtername_list_s_t {long size; / * number of event filter names * / [size_is (size)] / * event filter name in group * / ems_string_t filter_name [];} ems_filtername_list_t;

Event Filter List The structure of the event filter list contains a list of size filters. typedef struct ems_filter_list_s_t {long size; / * number of event filters * / [size_is (size)] ems_filter_t * filter []; / * event filter list * /} ems_filter_list_t;

-EMS Use Process-Data Structure-Use Process The data structure of the use process defines the EMS use process. Each usage process has a name entered in the CDS, a host name of the host on which the usage process is running, and a uuid unique to the usage process. typedef struct ems_consumer_s_t {[string] char * name; / * DCE name of used process * / ems_netname_t * hostname; / * DCE host name of used process * / uuid_t uuid; / * uuid * /} ems_consumer_t; of used process

Used Process List The used process list structure contains a list of sizes of used process entries. typedef struct ems_consumer_list_s_t {long size; / * number of used process entries * / [size_is (size)] ems_consumer_t consumer []; / * used process information * /} ems_consumer_list_t;

EMS Server Data Structure Attribute List The attribute list data structure defines a list of server attributes. Each attribute is a value maintained by the emsd server and its attribute list can be used to query and set these values. typedef struct ems_attrlist_s_t {long size; / * number of attributes * / [size_is (size)] ems_attrbute__t attr []; / * event format attribute * /} ems_attrlist_t;

Application Program Interface (API) The following is a given representative routine of EMSAPI that implements the RPC interface. -Register to EMS #include <dce / ems.h> void ems_register (ems_netname_t * hostname, / * DCE host name * / ems_handle_t * handle, / * EMS handle * / error_status_t * status); / * mgmt request status * / Description: This routine registers with EMS and gets the EMS binding handle. This routine can be utilized by management applications that will use the EMS management mechanism, or by event provisioning processes that are trying to add new event types. Authorization request. No reply
(None). Parameters: Input hostname: The name of the DCE host machine on which emsd is running. If the host name is null, it is considered the local host. Output handle: Returns an EMS handle for use in later calls to EMS routines. status: Returns a status code from this routine indicating whether the routine completed normally and why it did not complete normally. The possible status codes and their meanings are: error_status_ok: Completed normally. ems_s_no_memory: EMS handle cannot be allocated. ems_s_unsupported_namesevice: Hostnames contain unsupported name services.

Transmission by remote provision process void rsupplier_transmit ([in] handle_t hadle, [in] ems_event_t * event, [out] error_status_t * status); Description: Transmits the event to the EMS. Authorization request. (w) on /.;/hosts/<hostname>/ems-server/event-t
ypes / <type_name> Parameter: Input handle: EMS binding handle event: Event to be transmitted Output status Rturns: Status code from this routine that indicates whether the routine completed successfully and why it did not complete successfully . The possible status codes and their meanings are: error_status_ok: Completed normally. ems_s_no_memory: The EMS server received a memory allocation error.

Add event filter to remote use process group void rconsumer_add_filter_to_group ([in] handle_t handle, [in] cons_context_t ch, [in] ems_filtername_list_t * filter_list, [out] error_status_t * status); Description: Use process Add event filters from the filter list (filter_list) to the event filter group. The using process is identified by its uuid. Authorization request. (w) on /.:/ host / <hostname> / ems-server / consumers Parameter: Input handle: EMS binding handle ch: Context output of used process returned from remote used process register status Rturns: Routine completed normally This routine returns a status code indicating whether or not it has completed and the reason why it did not complete normally. The possible status codes and their meanings are: error_status_ok: Completed normally.

-Transmission by EMS void ems_transmit ([in] handle_t handle, [in] ems_event_t * event, [out] error_status_t * status); Description: Transmits an event from the EMS to the used process. Parameters: Input handle: Used process binding handle event: Event output status Rturns: Returns a status code from this routine that indicates whether the routine completed successfully and why it did not complete successfully. The possible status codes and their meanings are: error_status_ok: Completed normally.

One of the preferred embodiments of the event management service according to the present invention is an instruction set in a code module resident in the random access memory of the computer. This instruction set may be stored in another computer memory until required by the computer. For example, it is a removable memory such as a hard disk drive, an optical disk (which is appropriately used in a CDROM), or a floppy disk (which is appropriately used in a floppy disk drive). Moreover, while the various methods described above are conveniently implemented on a general-purpose computer that is selectively activated or recognized by software, one of ordinary skill in the art would appreciate that
It will be appreciated that these methods can be implemented in hardware built to perform the necessary steps, in firmware, or even on more specialized devices.

FIG. 8 shows a block diagram of a typical computer system in which the present invention may be implemented. The system unit 21 comprises one or a plurality of system buses 131 in which various components are combined to realize communication between these components. Microprocessor 1
32 is connected to the system bus 131, and is also a read-only memory (R
OM) 133 and random access memory (RA
M) supported by 134. Microprocessors include Intel, including 386 or 486 microprocessors
Family of microprocessors, other microprocessors such as the Motorola family (68000, 68020 or 68030) or IBM, Hewlett Packard, Su
It may be a RISC microprocessor manufactured by Intel, Motorola, et al.

ROM 133 contains the Basic Input / Output System (BIOS), among other code, which controls the operation of basic hardware such as dialogs and disk drives and keyboards. RAM 134 is the main memory into which the operating system and application programs are loaded. Memory management chip 135
Is connected to the system bus 131 and controls direct memory access operations including data transmission between the RAM 134 and the hard disk drive 136 and floppy disk drive 137. CDROM
142 is also connected to the system bus 131 and is used to store large amounts of data such as multimedia programs or large databases.

To this system bus 131, various I
The / O control device is also connected. For example, keyboard controller 138, mouse controller 139, video controller 140, and audio controller 141. Keyboard controller 138 provides a hardware interface for keyboard 122. Mouse controller 139 provides a hardware interface for mouse 123. Video controller 140 is a hardware interface for display device 124. The audio control device 141 includes a speaker 125a.
And 125b are hardware interfaces. I / O controller 150, such as a token ring adapter, enables communication over network 156 to other similarly constructed data processing systems.

As shown in FIG. 8, the RAM 134 has an event management service (E) including a filter mechanism and a queue mechanism.
MS) 22 is supported. The thread service is shown at 158 and includes a plurality of thread process control routines 160a-160n.

Although the present invention has been described above with reference to a preferred embodiment for a particular operating system and network environment (OSF DCE), those skilled in the art will appreciate other different operating system and network environments.
It will be appreciated that the invention can be implemented with changes in architecture. The present invention is based on the OSF DCE
It should not be construed as limited to the architecture, and therefore, more generally, the present invention should be broadly construed as encompassing any network environment, whether conventional or object oriented.

In summary, the following matters will be disclosed regarding the configuration of the present invention.

(1) A method of managing events in a distributed computing environment in which a distributed resource is accessible by a user and one or more event providing processes generate an event used by one or more event using processes. A step of defining by the event using process which one of a plurality of events generated by the one or more event providing processes should be transmitted to the one event using process, and a predetermined remote procedure A step of determining by an event management service whether an event related to a call (RPC) is one of the plurality of events transmitted to the event use process, and the event related to the predetermined RPC is Of the plurality of events transmitted to the event usage process If it is One, method of managing events in a distributed computing environment, including the steps of providing the said event from the event management service to the event consumer. (2) The event management method in the distributed computing environment according to (1), wherein the predetermined RPC is issued from an event recording service. (3) The event management method in the distributed computing environment according to (1), wherein the event is provided from the event management service to the event using process using a remote procedure call. (4) The event management method in the distributed computing environment according to (1), including the step of registering the event use process in the event management service. (5) The event management method in the distributed computing environment according to (4) above, which includes a step of authenticating the event use process with a security service. (6) The event using process selects which event to receive from the one or more event providing processes by defining an event filter group having one or more event filters. An event management method in the distributed computing environment described in. (7) The event management method in the distributed computing environment according to (6), wherein each event filter of the event filter group includes a set of filter expressions. (8) The event management method in the distributed computing environment according to (7), wherein the filters pass the event when all the filter expressions in one filter meet a predetermined condition. (9) The event management method in the distributed computing environment according to (7), wherein when any one of the filter expressions in one filter does not meet a predetermined condition, the filter does not pass the event. (10) The event management method in the distributed computing environment according to (9), in which the next filter in the event filter group is evaluated when the filter does not pass the event. (11) A method of managing an event in a distributed computing environment in which a distributed resource is accessible by a user and one or more event providing processes generate an event used by one or more event using processes. , At least one event provision process is 1
Which of the plurality of events generated by the one or more event providing processes should be transmitted to the event using process when the event is distributed on a machine away from the one event using process; In response to the steps defined by each of the consuming processes and the remote procedure call from the event recording service, the event associated with each remote procedure call should be transmitted to any of the plurality of event consuming processes. The event management service determines if the event is an event, and if the event associated with the remote procedure call is to be transmitted to any of the plurality of event consuming processes, the remote procedure call. By using a call, METHOD event management from Preparative the event management service to all event consumer that requested the transmission of the distributed computing environment comprising the steps of providing the event. (12) The event management method in the distributed computing environment according to (11) above, including a step of authenticating at least one event using process. (13) At least one event use process is 1
Alternatively, in the distributed computing environment according to (11), which event is to be transmitted from the one or more event providing processes by defining an event filter group having a plurality of event filters. Event management method. (14) The event management method in the distributed computing environment according to (11) above, wherein at least one event providing process does not recognize the event using process registered in the event management service to receive an event of the event providing process. (15) The event management method in the distributed computing environment according to (11) above, which includes a step of storing the event in an event log. (16) A computer network provided with a distributed computing environment in which distributed resources can be accessed by users, and an event channel for passing an event from one or more event providing processes to one or more event using processes And an event management service means for establishing an event, and a remote procedure call for providing an event from the one or more event providing processes to the event management service means, and from the event management service means to the one or more event using processes. A computer network having means. (17) The computer according to (16) above, which has access control means for authenticating an event using process in order to receive an event from the event management service means.
network. (18) The computer network according to (16) above, which has an event recording service means for giving an event from the one or more event using processes to the event management service means. (19) A computer system used in a distributed computing environment, comprising: a processor, an operating system executed on the processor, and one or more event providing processes, and one or more event using processes. Event management service (EMS) program means executed by the operating system to establish an event channel for passing, and from the one or more event providing processes to the EMS program means and from the EMS program means A remote procedure call program means executed by the operating system to provide events to the one or more event consuming processes. (20) A computer-readable storage medium having a substrate, and computer program data encoded on the substrate of the computer-readable storage medium for causing a computer to manage event data. A remote procedure call for establishing an event channel through which the computer program data asynchronously passes events from one or more providing processes to one or more using processes
An article of manufacture that has an Event Management Service (EMS) that responds to (RPC).

[Brief description of drawings]

FIG. 1 shows a computer network in which the present invention is implemented.

FIG. 2 illustrates a representative distributed computing domain with connected machines in which the present invention is implemented.

FIG. 3 is a block diagram illustrating an exemplary embodiment of the event management service of the present invention within the context of an existing DCE subsystem.

FIG. 4 is a block diagram illustrating a method of configuring an event filter according to the present invention.

FIG. 5 is a flow chart illustrating an event-using process parsing routine of a filter mechanism according to the present invention.

FIG. 6 is a block diagram showing a queue configuration of an event management service.

FIG. 7 is a flow chart showing the event queue management routine of the present invention.

FIG. 8 is a block diagram of a computer system in which the event management service of the present invention is implemented.

[Explanation of symbols]

 3 Network 10 Processing System 12 DCE Cell 14 Server 16 DCE Client 18 Security Service 20 Event Recording Service 22 Event Management Service (EMS) 24 Event Providing Process 26 Event Using Process 28 DCE Audit 30 DCESVC 31 RPC 32 EMSAPI Layer 34 Audit Trail 36 SVC log file 38 Console 40 Process database used 42 Event log file 44 Event format database 45 Filter 46 Event filter database 47 Queue

 ─────────────────────────────────────────────────── —————————————————————————————————————— Inventor Gregory Alan Wilson, Walnut Clay 3917, Austin, TX 78731, USA

Claims (20)

[Claims] 1. A method for managing events in a distributed computing environment in which distributed resources are accessible by a user and one or more event providing processes generate events used by one or more event consuming processes. And a step of defining by the event using process which one of the plurality of events generated by the one or more event providing processes should be transmitted to the one event using process, and a predetermined remote procedure call A step of determining whether an event related to (RPC) is one of the plurality of events transmitted to the event use process, the event related to the predetermined RPC is One of the plurality of events transmitted to the event consuming process In some cases, method of managing events in a distributed computing environment, including the steps of providing the said event from the event management service to the event consumer. 2. The method of managing events in a distributed computing environment according to claim 1, wherein the predetermined RPC is issued from an event recording service. 3. The event management method in a distributed computing environment according to claim 1, wherein the event is provided from the event management service to the event using process using a remote procedure call. 4. The event management method in a distributed computing environment according to claim 1, further comprising the step of registering the event use process in the event management service. 5. The event management method in a distributed computing environment according to claim 4, further comprising the step of authenticating the event use process with a security service. 6. The event consuming process selects which event is received from the one or more event providing processes by defining an event filter group having one or more event filters. 1. An event management method in the distributed computing environment according to 1. 7. The event filter of the event filter group includes a set of filter expressions.
An event management method in the distributed computing environment described in. 8. The event management method in a distributed computing environment according to claim 7, wherein when all the filter expressions in one filter satisfy a predetermined condition, the filter passes the event. 9. The event management method according to claim 7, wherein the filter does not pass the event when any one of the filter expressions in one filter does not meet a predetermined condition. 10. The method of managing events in a distributed computing environment of claim 9, wherein the next filter in the event filter group is evaluated if the filter does not pass the event. 11. A distributed resource is accessible by a user and one or more event providing processes are
Or a method of managing events in a distributed computing environment that generates events used by multiple event-using processes, wherein at least one event-providing process is located on a machine remote from the one event-using process In which each of the plurality of event using processes defines which of the plurality of events generated by the one or more event providing processes should be transmitted to the event using process; Remote procedure
In response to the call, the event management service determines whether the event associated with each remote procedure call is an event to be transmitted to any of the plurality of event consuming processes; If the event related to the procedure call is to be transmitted to any of the plurality of event consuming processes, a remote procedure call is used so that all the event consuming processes that have requested the transmission of the event. And providing the event from the event management service. 12. The method of managing events in a distributed computing environment of claim 11 including the step of authenticating at least one event usage process. 13. Which event should be transmitted from at least one event providing process by at least one event consuming process by defining an event filter group having one or more event filters. The event management method in the distributed computing environment according to claim 11, wherein 14. The method of managing events in a distributed computing environment according to claim 11, wherein at least one event providing process does not recognize the event using process registered in the event management service to receive an event of the event providing process. . 15. The method of managing events in a distributed computing environment according to claim 11, including the step of storing the events in an event log. 16. A computer network provided with a distributed computing environment in which distributed resources can be accessed by users, and an event for passing an event from one or more event providing processes to one or more event using processes.・
An event management service means for establishing a channel, and a remote procedure for providing an event from the one or more event providing processes to the event management service means, and from the event management service means to the one or more event using processes. A computer network having call means. 17. The computer network of claim 16 including access control means for authenticating an event consuming process to receive events from said event management service means. 18. The computer network of claim 16, further comprising event recording service means for providing events from the one or more event usage processes to the event management service means. 19. A computer system used in a distributed computing environment, comprising: a processor, an operating system executed on the processor, and one or more event providing processes to one or more event using processes. Events for passing events
Event management service (EMS) program means executed by the operating system to establish a channel; and the EMS from the one or more event providing processes.
A remote procedure call program means executed by the operating system to provide events to and from the EMS program means to the one or more event consuming processes. 20. A computer readable storage medium comprising a substrate and computer program data encoded on the substrate of the computer readable storage medium for causing a computer to manage event data. An event having the computer program data responding to a remote procedure call (RPC) to establish an event channel for asynchronously passing an event from one or more providing processes to one or more using processes. An article of manufacture with a management service (EMS).