 |
Description  |
|
|
CROSS REFERENCE TO RELATED APPLICATIONS
This application is related to co-pending U.S. patent application Ser. Nos.
07/720,259 (Atty. Docket No. 09447/0065) entitled COMPUTER SYSTEM MANAGER,
07/719,240 (Atty. Docket No. 09447/0068) entitled INNATE BUS MONITOR FOR
COMPUTER SYSTEM MANAGER, 07/719,243 (Atty. Docket No. 09447/0070) entitled
REMOTE CONSOLE EMULATOR FOR COMPUTER SYSTEM MANAGER, and 07/719,394 (Atty.
Docket No. 09447/0074) entitled POWER SUPPLY FOR COMPUTER SYSTEM MANAGER,
all of which were filed on even date herewith, assigned to the assignee of
the present invention and hereby incorporated by reference as if
reproduced in their entirety.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to hardware manageability support systems in
network environments. More particularly, the present invention relates to
systems that deliver alerts in local area network, wide area network,
enterprise network environments, and the like.
2. Description of Related Art
In its simplest form, data communication takes place between two devices
that are directly connect by some form of point-to-point transmission
medium. However, in certain instances it is impractical for two devices or
components to be directly, point-to-point connected. An example of such an
instance is when there are a set of devices, each of which may require a
link to many of the others at various ties. This problem has been
heretofore addressed by attaching multiple devices to a communication
network. Each device, or station, is attached to a network node. The set
of nodes to which stations attach is the boundary of a communication
network that is capable of transferring data between pairs of stations.
Communication networks may be categorized based on the architecture and the
techniques used to transfer data. Two major different types of
communication networks are switched networks and broadcast networks.
Switched networks involve transfer of data from a source to a destination
through a series of intermediate nodes that provide a switching facility
to move the data from node to node. Examples of switched communication
networks include circuit switched networks, message switched networks, and
packet switched networks. Broadcast networks involve no intermediate
switching nodes; each station has a transmitter/receiver that allows
communication over a common medium, so that a transmission from any one
station is broadcast to and received by all other stations in the network.
Examples of broadcast communication networks include packet radio
networks, satellite networks, and local networks.
The present invention is especially designed to operate in
packet-broadcasting networks such as local area networks ("LANS"), wide
area networks ("WANS"), enterprise network environments ("enterprise
environments") and the like. As those skilled in the art should
appreciate, the terms LANs, WANs and enterprise environments have
constantly evolving meanings. In general, however, a LAN is a
general-purpose local network supporting personal computers ("PCs"), a WAN
is a general-purpose network supporting a wide area relative to a LAN
(usually because it passes through a public switched telephone network at
least in part), and an enterprise environment is a general-purpose network
supporting mini computers and mainframes. LANs, WANs, and enterprise
environments can also support any number of other peripherals (e.g.,
printers), and quite often carry voice, video, and graphics as well as
data. To assist those presently familiar with and skilled in the art in
understanding what is meant by the various network terms used herein, the
term IAN as used herein encompasses Novell NetWare, SOD UNIX, Banyan
VINES, Microsoft LAN Manager, and the like. Further, the term enterprise
environment as used herein encompasses IBM SNA Management Services, DEC
Enterprise Management Architecture, AT&T Unified Network Management
Architecture, SNMP-Based Management Stations, Hewlett-Packard's OPENVIEW,
and the like.
As will become apparent from the detailed description of the invention
section below, the present invention may be viewed as an alert delivery
system servicing a number of system components that supplement
conventional network components to enhance management functions. These
conventional network components include network management agents and
network manager consoles. Underlying both of these components is a network
operating system ("OS"). The OS is software, firmware, or hardware that
controls the execution of processes in a system. Services provided by an
OS can include resource allocation scheduling, input/output control, and
data management. The theory b an OS is to provide support in a single spot
so that each program does not need to be concerned with controlling
hardware. A network management agent is a part of the OS. The network
management agent is the central collection point for a particular server's
network management information. Typically, the information is stored in an
object oriented format for use by "network manager" software that runs on
the network such as a remote console or IBM's Netview. A network manager
console, the second component mentioned above, is generally provided as an
optional component piece of an OS network management package. Typically,
the network manager console allows one central point of monitoring and
control for a network administrator's domain of interest. From the console
the administrator can view and control the objects registered with the
network management agent. The network manager console represents one type
of manager software connected over the internet (i.e., interconnected
networks) with the network management agent. Other network manager
software can also make use of the information in the management agent.
Two other concepts that should be fully understood to appreciate the
present invention are "in-band" and "out-of-band". The term in-band is
used herein to refer to signals sent via network management software
residing locally on the server. In simple terms, in-band refers to use of
the wire that comes with a network. The term out-of-band refers to
communications that leave the computer system manager (discussed at length
in the related case so titled) via an asynchronous port. Although the
out-of-band concept will be discussed in much greater detail below, it
should be appreciated that asynchronous transmission is involved, that is,
transmission in which each information character is individually
synchronized, usually by the use of start elements and stop elements.
Although network managers have heretofore been able to provide an
acceptable level of network management, the constant improvements being
made in networks betray the fact that many improvements can still yet be
made. For ale, it is important that the system software or the network
manager be notified about significant or potentially significant events
occurring or about to occur in the network. Alerts perform such a
function. However, prior art alert delivery systems have a number of
shortcomings and defects. For example, prior art alert delivery systems
are limited as to the number and types of alert receivers, the alert
transmission media that can be used, capability for retrying, and
effectiveness of the underlying algorithm, to name just a few.
SUMMARY OF THE INVENTION
The present invention overcomes the shortcomings and deficiencies of the
prior art by providing an in-band/out-of-band alert delivery system for a
computer system manager including an alert log which maintains a record of
alerts to be delivered and the status of those alerts, an alert manager
for making a first attempt to deliver each alert, and a retry manager for
making subsequent attempts to deliver alerts as becomes necessary and
appropriate.
In embodiments of the alert delivery system of the present invention there
may be a bus master interface manager for making in-band alert deliveries
and a communications manager for making out-of-band alert deliveries.
Telephone numbers may be provided to the communications manager by an
alert destination list. Out-of-band alert deliveries may be made via a
modem, a universal asynchronous receiver transmitter, or the like.
According to the teachings of the present invention, out-of-band deliveries
may be made to pagers, telephones and remote consoles via asynchronous
lines. The underlying algorithm in embodiments of the present invention is
designed for maximum success rate, via exhaustive retrying, in delivering
alerts.
Accordingly, it is an object of the present invention to provide an
improved alert delivery system for incorporation into data communication
network environments.
Another object of the present invention is to provide an
in-band/out-of-band alert delivery system with a persistent retrying
capability.
Yet another object of the present invention is to provide an
in-band/out-of-band alert delivery system for a computer system manager
that is able to deliver alerts to many different types of alert receivers
including pagers, telephones and remote consoles.
Still yet another object of the present invention is to provide an
in-band/out-of-band alert delivery system for a computer system manager
which operates based upon a very effective "underlying" algorithm.
BRIEF DESCRIPTION OF THE DRAWINGS
Other objects, advantages, and novel features of the present invention will
become apparent from the following detailed description of the invention
when considered in conjunction with the accompanying drawings wherein:
FIG. 1 is a block diagram of a computer network having an EISA based
computer system and a computer system manager board installed therein;
FIG. 2 is a block diagram of portions of the EISA system board of FIG. 1
which illustrates information flow to and from, as well as within, the
board;
FIG. 3 is a flow diagram illustrating decisional processes in an embodiment
of the alert delivery stem according to the teachings of the present
invention; and
FIG. 4 is a flow chart depicting the logic employed in an embodiment of the
present invention.
DETAILED DESCRIPTION OF THE INVENTION
As discussed in detail in other of the related cases referenced above, the
management system in which the alert delivery system discussed herein may
be incorporated is in one embodiment a stem of hardware and software
components that provide enhanced hardware management capabilities for
EISA-based systems. In this embodiment, the management system is designed
to be incorporated into LANS, WANS, enterprise environments, and the like.
In greater detail, and as also discussed in greater detail in other of the
related cases referenced above, in various embodiments the system of
hardware and software components can include a 32-Bit intelligent Bus
Master EISA board and firmware, an OS device driver, a remote asynchronous
driver, a remote asynchronous console program, and configuration support.
These components are coupled with a network management agent and a network
manager console, both of which are discussed in the description of related
art section above, so as to provide enhanced hardware management
capabilities. Because these various components are discussed in detail in
other of the related cases referenced above, they will not be discussed in
detail herein. However, it may be noted that the 32-Bit intelligent Bus
Master EISA board and firmware provides the hardware and firmware support
for alerting controlling and monitoring functions, which functions are
discussed further below as well as in the related cases. The OS device
driver, a component which is necessary for each supported operating
system, is a mechanism for providing two-way communication between a
network management agent running on the operating system and the system
manager hardware. The remote asynchronous driver is a mechanism for
enabling access to the 32-Bit intelligent Bus Mater EISA board mentioned
above through an asynchronous connection. Further, the remote asynchronous
driver enables remote console software operation as discussed in other of
the related cases referenced above. The remote asynchronous console
program is a software component that executes on any PC connected to a
modem or asynchronous interface of the system manager. This software
interacts with the 32-Bit intelligent Bus Master EISA board by calling a
resident asynchronous driver API that has a protocol link with an
asynchronous or modem port of the system manager. The API allows remote
control (e.g., rebooting), access to stored data, and a remote console
function. The remote console emulation (text mode only) is independent of
the system and OS allowing significant functionality without the support
of the network or operating system. Remote console emulation is the
express subject matter of related U.S. patent application Ser. No.
07/719,243 (Atty. Docket No. 09447/0070) incorporated herein. All
out-of-band accesses to the modem or asynchronous interface of the system
manager can be made to require security checks before access is allowed.
The final component of the management system listed above, configuration
support, involves configuring the 32-Bit intelligent Bus ester EISA board
into an EISA system using an EISA configuration utility with corresponding
and appropriate files. Default settings for the cards' configurable
options can be stored in the host EISA nonvolatile memory. The management
system may also require runtime configuration, possibly remotely, to set
thresholds for alerting temperatures and other specific configuration
options discussed further below. These configuration commands can be made
available at various points throughout the management.
The system manager described in the immediately preceding paragraph can
perform a number of functions. Four of the many functions it can perform
are alerting, monitoring, remote management, and histories. Although
alerting will be discussed in great detail below, it may be noted at this
point that the management system discussed in general in the various
related applications is able to dial a preconfigured telephone number and
report important events, that is, to alert. Alerting can be accomplished
by any of the following methods: by sending a protocol message over a
switch telephone connection or through a direct serial connection, by
dialing a phone number associated with a pager, and/or by dialing a phone
number associated with a person and leaving a synthesized voice message.
The foregoing list is not exhaustive as will be discussed further below.
The second function mentioned above, monitoring, is the ability of the
system to monitor system components and parameters. Simply, this means
that if a component experiences a failure, or exhibits characteristics
that indicate that it might have a failure, portions of the management
system can detect this and report it to a system software entity that can,
in turn, report the information to a network administrator so that
corrective action can be taken. Three examples of a multitude of system
components and parameters that can be monitored by the management system
include system power, internal temperature, and memory refresh and parity.
In all three of these examples alerts or portions of the alerting system
may be involved. For example, if the system detects a drop in system power
(i.e., below some predefined threshold), it may report the event as an
alert. Likewise, if the temperature rises above a predetermined threshold,
the system may issue an alert. Still further, if the system detects the
absence of memory refresh signals or a memory parity error, it may issue
an alert. In general, associated with the various system objects within
the system may be a capability to provide generic thresholding and
alerting. That is, if same counter or value within a system object exceeds
a threshold, the system can automatically send an alert either locally or
through one or both of its serial ports.
The remote management function refers to the ability to provide remote
management of a server. As previously discussed, alerting is a function
that may be provided remotely. Additionally, the management system can, in
some embodiments, recognize a small list of tones over its modem
connection. These tones may be used to issue commands. For example, when
the management system issues an alert over a modem using voice synthesis,
a touchtone command may be used to "acknowledge" the alert.
Based upon the foregoing, it should be readily apparent to those skilled in
the art that the management system discussed in this and the various
related cases has an alert delivery system, which alert delivery system is
not only important independently, but which system is also important
because its capabilities and operations management permeate and influence
various other components throughout the system. With that in mind, the
alert system may now be discussed in detail.
Referring initially to FIG. 1, an extended industry standard architecture
("EISA") based computer network 10 is shown therein. The computer network
10 includes an EISA server 12 and an EISA board/bus 13. Installed on the
EISA system board 13 are a plurality of cards including a system manager
22, an EISA network interface adapter 24, and an intelligent disk array
controller device 26. Also running on the EISA server 12 is a network
operating system 14, preferably one which includes a network management
agent 15. Two-way communication between the system manager 22 and the
network management agent 15 is provided by a system manager device driver
16. In a similar manner, a network interface device driver 18
interconnects the network management agent 14 and the EISA network
interface adapter 24, and an intelligent disk array controller device
driver 20 interconnects the network operating system 14 and the
intelligent disk array controller device 26. Also shown in FIG. 1 are
three terminals connected to the overall network 10. Terminal 30 is
connected to the EISA network interface adapter 24 via EISA bus 28.
Terminal or network manager console 34 is connected to the system manager
22 via an asynchronous (out-of-band) line 40. Finally, terminal or network
manager console 36 is connected to the network operating system 14 via
inter-network communication (in-band) line 42. The network manager console
34 may operate in either a remote or an out-of-band local mode. The
network manager console 36 operates only in a local mode.
Alerts are used to notify system software or a network administrator about
important events. For example, if the host experiences a power failure,
the enhancement or EISA system board 13 detects the loss of power and
sends an alert. As heretofore mentioned, there are two types of alerting:
in-band or out-of-band. In-band alerts are sent to software residing
locally on the server 12 with the enhancement board 13. If an alert is to
be delivered in and out-of-band, one or the other (typically the in-band
alert) may be delivered first. Out-of-band alerts are sent through the
EISA system board's asynchronous port (designated by reference numeral 44
in FIG. 1) to a remote application. These alerts include alerts sent to a
remote console, a phone, or a pager.
Referring now to FIG. 2, further details regarding the alert system are set
forth therein. In that Figure, it may be seen that asynchronous lines 46,
48, 50 may exist between the EISA system board 13 and a pager 52, a
telephone 54 and a console 56, respectively. Although further details will
be provided below, it may be noted in FIG. 2 that from a functional point
of view, the EISA system board 13 includes an alert management controller
or alert manager 58 which can trigger operation of a modem 60 with voice
capability 62 as appropriate based upon comparison of values obtained from
a bus monitor 62 and threshold values in an object data file 64.
Referring now to FIG. 3, there is shown a flow diagram illustrating
decisional processes in embodiments of the alert delivery system according
to the teachings of the present invention. Operation of the alert delivery
system according to the teachings of the present invention begins with
generation of an alert and delivery of that alert to the alert manager
(discussed above with respect to FIG. 2). Promptly upon receiving a new
alert, the alert manager 58 has the fact that the alert was received
entere | | |
