| Rfc | 0712 |
|---|
| Title | Distributed Capability Computing System (DCCS) |
|---|
| Author | J.E. Donnelley |
|---|
| Date | February 1976 |
|---|
| Format: | TXT, PDF, HTML |
|---|
| Status: | UNKNOWN |
|---|
|
Network Working Group J.E. Donnelley
Request for Comments: 712 Lawrence Livermore Laboratory
February 1976
A Distributed Capability Computing System (DCCS)
This paper was prepared for submission to the international
Conference on Computer Communication, ICCC-76, August 3, 1976,
Toronto, Canada.
This is a preprint of a paper intended for publication in a journal
of proceedings. Since changes may be made before publication, this
preprint is made available with the understanding that it will not be
cited without the permission of the author.
The work reported in this paper was supported in part under contract
#EPA-IAG-D5-E681-DB with the Environmental Protection Agency and in
part under contract #[RA] 76-12 with the Department Of
Transportation. The report was prepared for the U.S. Energy Research
and Development Agency under contract #W-7405-Eng-48.
A Distributed Capability Computing System (DCCS)
This paper describes a distributed computing system. The first
portion introduces an idealized operating system called CCS
(Capability Computing System). In the second portion, the DCCS
protocols are defined and the processes necessary to support the DCCS
on a CCS are described. The remainder of the paper discusses
utilizing the DCCS protocol in a computer network involving
heterogeneous systems and presents some applications. The
applications presented are to optimally solve the single copy problem
for distributed data access and to construct a transparent network
resource optimization mechanism.
The Capability Computing System (CCS)
The CCS, though not exactly like any existing operating system, is
much like some of the existing capability list (C-list) operating
systems described in the literature [1-7]. Many of the features of
the CCS come from a proposed modification to the RATS operating
system [1-3].
In the documentation for most computer systems there are many
references to different types of objects. Typical objects discussed
are: files, processes, jobs, accounts, semaphores, tasks, words,
devices, forks, events, etc. etc.. One of the intents of C-list
systems is to provide a uniform method of access to all such objects.
Having all CCS objects accessed through a uniform mechanism allow
DCCS to be implemented in a type independent manner.
The CCS is a multiprocessing system supporting an active element
called a process. For most purposes, the reader's intuitive notion
of what a process is should suffice. A process is capable of
executing instructions like those in commercially available
computers. It has a memory area associated with it and has some
status indicators like "RUN" and "WAIT". In C-list systems, however,
a process also has a capability list (C-list). This list is an area
in which pointers to the objects that the process is allowed to
access are maintained. These pointers are protected by the system.
The process itself is only allowed to use its C-list as a source of
capabilities to access and as a repository for capabilities that it
has been granted. Figure 1 diagrams some typical processes that are
discussed later. In the diagrams, the left half of a process box is
the C-list and the right half is the memory.
The key to the uniform access method in the CCS is the invocation
mechanism. This is the mechanism by which a process makes a request
on a capability in its C-list. An invocation is closely analogous to
a subroutine call on most computer systems. When a request is made,
the invoking process passes some parameters to a service routine and
receives some parameters in return.
There are, however, several major differences between the invocation
mechanism and the usual subroutine calling mechanisms. The first
difference is that the service routine called is generally not in the
process's memory space. The service routine is pointed to by the
protected capability and can be implemented in hardware, microcode,
system kernel code, in another arbitrary process, or, as we shall see
in the DCCS, in another computer system. In Fig. 1. for example, the
serving process is servicing on invocation on the semaphore
requestor.
A second difference is that, when invoking a capability, other
capabilities can be passed and returned along with strictly data
parameters. In the DCCS, capabilities and data can also be passed
through a communication network.
The final important distinction of the invocation mechanism can best
be illustrated by considering the analogy to the outside teller
windows often seen at banks. These windows usually contain a drawer
that can be opened by the customer and teller are not both. Except
for this drawer, the customer and teller are physically isolated. In
the case of the invocation mechanism, the invoking process explicitly
passes certain capabilities and information to the service routine
and designated C-list locations and memory areas for the return
parameters. Except for these parameters, the invoking process and
the serving routine are isolated. In the DCCS, this protection
mechanism is extended throughout a network of systems.
In the CCS, invoking a capability is the only way that a process can
pass or receive information or capabilities. All of what are often
referred to as system calls on a typical operating system are
invocations on appropriate capabilities in the CCS. A CCs C-list
envelopes its process. This fact is needed in order to transparently
move processes as described in the second application on network
optimization (page 23).
CCS Capabilities
To build the DCCS, we will assume certain primitive capabilities in
the CCS. The invocations below are represented for simplicity rather
than for efficiency or practicality. In practice, capabilities
generally have more highly optimized invocations with various error
returns, etc.. To characterize a capability, it suffices to describe
what it returns as a function of what it is passed. In the notation
used below, the passed parameter list is followed by a ">" and then
the returned parameter list. In each parameter list the data
parameters are followed by a "" and then the capability parameters.
1. File Capability
a. "Read", index; > data;
"Read" the data at the specified index. "Read" and the index
are passed. Data is returned.
b. "Write", index, data; > ;
Write the data into the area at the specified index. "Write",
the index, and the data are passed. Nothing is returned.
2. Directory Capability
a. "Take", index; > ; capability
"Take" the capability from the specified index in the
directory. "Take" and the index are passed. The capability is
returned.
b. "Give", index; capability> ;
"Give" the capability to the directory at the index specified.
"Give" and the index are passed information. The capability is
also passed. Nothing is returned.
c. "Find"; capability> result, index;
A directory, like a process C-list, is a repository for
capabilities. The first two invocations are analogous to the
two file invocations presented except that they involve
capability parameters moved between directory and C-list
instead of between file and memory. The last invocation
searches the directory for the passed capability. If an
identical capability is found, "Yes" and the smallest index of
such a capability are returned. Otherwise "No" and 0 are
returned.
3. Nil Capability
When a directory is initially created, it contains only nil
capabilities. Nil always returns "Empty".
4. Process Capability
a. "Read", index; > data;
b. "Write", index, data; > ;
c. "Take", index; > ; capability
d. "Give", index; capability> ;
e. "Find"; capability> result, index;
f. "Start"; > ;
g. "Stop"; > ;
The a. and b. invocations go to the process's memory space. C., d.,
and e. go to its C-list. F. and g. start and stop process execution.
The CCS Extension Mechanism
There is one more basic capability mechanism needed for the CCS
implementation of the DCCS. This mechanism allows processes to set
themselves up to create new capabilities that they can service. Such
mechanisms differ widely on existing C-list systems. A workable
mechanism is described. Another primitive capability is needed to
start things off:
5. Server Capability
a. "Create requestor", requestor number; > ; requestor
b. "My requestor?"; capability> answer, requestor number;
c. "Wait"; > reason, requestor number, PD; request
Two capabilities were introduced above besides the server, the
requestor and request capabilities. These capabilities will be
described as the invocations on a server are described.
The first invocation creates and returns a requestor capability. The
number that is passed is associated with the requestor. The
requestor capability is the new capability being created. Any sort
of invocation can be performed on a requestor. This is their whole
reason for existence. A process with a server capability can make a
requestor look like any kind of capability.
The "My requestor?" invocation can be used to determine if a
capability is a requestor on the invoked server, it returns either:
"Yes", requestor number; or "No",0;
The last invocation "Wait"s until something that requires the
server's attention happens. There are two important events that a
service routine needs to be notified about. If the last capability
to a requestor is overwritten so that the requestor cannot again be
invoked until a new one is created, the "wait" returns:
"Deleted", requestor number, 0; Nil
The last two parameters, 0 and Nil, are just filler for the returned
PD and request (see 5c). When a "wait" returns "Deleted", the
service routine can recycle any resources being used to service the
numbered requestor (e.g., the requestor number).
The most important event that causes a "wait" to return is when one
of the requestors for the server is invoked. In this case the server
returns:
"Invoked", requestor number, PD; request
The third parameter, labeled PD, stands for Parameter Descriptor. It
describes the number of each kind of parameter passing each way
during a requestor invocation. Specifically, it consists of four
numbers: Data bits passed, capabilities passed, data bits requested,
and capabilities requested.
The last parameter received, the request capability, is used by the
serving process to retrieve the passed parameters and to return the
requested parameters to the requesting process. Accordingly, it has
the following invocations:
6. Request Capability
a. "Read parameters"; > {The passed parameters
b. "Return", {The return parameters}> ;
The "Return" invocation has the additional effect of restarting the
requesting process.
One thing that should be noted about the server mechanism is that
invocations on a server's requestors are queued until the server is
"wait"ed upon. This is one reason that a request is given a separate
capability. The serving process can, if it chooses, give the request
to some other process for servicing, while it goes back and waits on
its server for more requests. The corresponding situation in the
outside bank window analogy would be the case where the teller gives
the request to someone else for service so that the teller can return
to waiting customers. The request capability points back to the
requesting process so that the return can be properly effected.
A sample service, that of the well known semaphore [8] service
routine keeps a table containing the semaphore values for each
semaphore that it is servicing. It also keeps a list of queued
requests that represent the processes that become hung in the
semaphore by "P"ing the semaphore when it has a value less than or
equal to zero. The invocations on a semaphore are:
7. Semaphore
a. "P"; > ;
b. "V"; > ;
A diagram and flow chart for the semaphore serving process is given
in Figures 1. and 2. The flow charts are given include most of the
basic capability invocations, but do not include detailed
descriptions of table searches. The table structure for the
semaphore service routine includes entries for each supported
semaphore. Each entry contains the semaphore value and a link into a
list of pointers to the requests hung in the semaphore (if any).
The most important feature of the server mechanism is that, by using
it, the functioning of any capability can be emulated.
This property, similar to the insertion property discussed in [9], is
the cornerstone of the DCCS. The basic idea of the emulation is to
have the server "wait" for requests and pass them on to the
capability being emulated. Such emulation of a single capability is
flow charted in Figure 3. The emulation flow charted is an overview
that doesn't handle all situations correctly. For example, a
capability may not return to invocations in the same order that they
are received. These situations also appear in the DCCS, so their
handling will be discussed there rather than here. It is important
to note that, except for delays due to processing and communication,
the emulation done in the DCCS is exact.
The DCCS Implementation
The DCCS will initially be described on a network of CCS systems. We
will assume that there exists a network capability:
8. Network Capability
a. "Input"; > Host no., message;
b. "Output", Host no., message > ;
It is assumed that the "Output" invocation returns immediately
after queuing the message for output and that the "input"
invocation waits until message is available.
For pedagogical purposes, the description of the DCCS will be broken
into two parts. First a brief overview of the important mechanisms
will be given. The overview will gloss over some important issues
that will be resolved individually in the more complete description
that follows the overview.
The intent of the DCCS is to allow capabilities on one host to be
referenced by processes on other hosts having the appropriate
capabilities. To do this, each host keeps a list of capabilities
that it supports for use by other hosts. Each host also supports a
server, which gives out requestors that are made to appear as if they
were the corresponding capability supported by the remote host. When
one of these emulated requestors is invoked, its parameters are
passed by the emulating host through the network to the supporting
host. The supporting host then sees to it that the proper capability
is invoked and passed the parameters. When the invoked parameters
are passed back through the network to the emulating host. The
emulating host then returns the return parameters to the requesting
process.
For example, let us take the "Read" request on a file diagrammed in
figure 4. When the emulated file (a requestor) is invoked, the
emulating process receives "invoke", requestor number, PD; request.
The requestor number that is returned is actually a descriptor
consisting of two numbers: Host number, capability number. These
descriptors are called Remote Capability Descriptors (RCDs). An RCD
identifies a host and a capability in the list of capabilities
supported by that host. After receiving a request, the emulating
process reads the parameters passed by the requesting process and
sends them along with the Parameters Descriptor to the remote host in
an "invoke" message.
When the remote host receives this information, it passes the
parameters to the supported file capability by invoking it and
specifies the proper return parameters as noted in the Parameter
Descriptor. When the invoked file return parameters, the returned
data is passed back through the network to the emulating host in a
"Return" message. The returned data is then returned to the
requesting process by performing a "Return" invocation on the request
capability initially received by the emulating host. When the
requesting process is awakened by the return, it will appear to it
exactly as if a local file had been invoked.
This works fine when the parameters being passed and returned consist
simply of information, but what happens when there are capabilities
involved? In this case the routines use the existing remote
capability access mechanism and pass the appropriate descriptor. As
an example, the "Take" invocation on a directory is diagrammed in
figure 5. The only essential difference is the fact that a
capability has to be returned. When the capability is returned by
the invoked directory (or whatever it really is), the supporting host
allocates a new slot in its supported capability list for the
capability and returns a new descriptor to the emulating host. When
the emulating host receives the descriptor, it creates a new
requestor with the returned descriptor as its requestor number and
returns the requestor to the invoking process. The requestor so
returned acts as the capability taken from the remotely accessed
directory and can be invoked exactly as if were the real capability.
One important thing to notice about this mechanism is that neither
the emulating host nor the supporting host need to have any idea what
kind of capabilities they are supporting. The mechanism is
independent of their type. Also important is the fact that neither
host need trust the other host with anything more than the
capabilities that it has been rightfully granted. Even the DCCS
processes themselves need only be trusted with the network
capabilities and with the supported capabilities. Finally, note that
no secret passwords which might be disclosed are needed for security.
The DCCS directly extends the CCS protection mechanisms,
A more complete description of the DCCS will now be given. To avoid
unnecessary complication, however, several issues such as error
indications, system restart and recovery, network malfunctions,
message size limitations, resource problems, etc. are not discussed.
These issues are not unique to the DCCS and their solutions are not
pertinent here.
As noted earlier, the complete DCCS must address several issues that
were glossed over in the initial overview. As these issues are
discussed, several message types are introduced beyond the "Invoke"
and "Return" messages discussed in the overview. The formats for all
the DCCS messages are summarized in figure 6.
A. Timing -
Invocations can take a very long time to complete. We saw an
example in the semaphore capability earlier. An even more graphic
example might be a clock capability that was requested to return
nothing AFTER 100 years had passed. Clearly we don't want to have
the emulating process wait until it receives a "Return" message
from the remote host before servicing more invocations.
What is done in the emulating host is to add the request
capability to a list of pending requests after sending the
"invoke" message to the supporting host (this is somewhat like the
semaphore example earlier). The emulator can then go back and
wait for more local requests.
There is a similar problem on the supporting side. We don't want
the process waiting on the network input capability to simply
invoke the supported capability and wait for return. What it must
do is to set up an invocation process to actually invoke the
supported capability so that pending network input can be promptly
serviced. The invoking process must then return the parameters
after it receives them.
These additional mechanisms add complication of multiple requests
active between hosts. These requests are identified by a Remote
Request Number (RRN). The RRN is an index into the list of
pending requests.
B. Loops -
If host A passes a capability to host B, and B is requested to
pass the requestor that is being used to emulate the capability
back to host A, should B simply add the requestor to its support
list and allow A to access it remotely? If it did, when the new
requestor was invoked on A, the parameters would be passed to B
where they would be passed to the requestor by the invoking
process. Invoking the requestor would cause the parameters to be
passed back through the network to A where the real capability
would finally be invoked. Then the return parameters would have
to go through the reverse procedure to get back A via B. This is
clearly not an optimal mechanism,
The solution to this problem makes use of the "My requestor?"
invocation on a server capability described in 5b. When B checks
a capability that is to be returned to A with the "My requestor?"
invocation and finds that the capability is one of its requestors
with a requestor number indicating that it is supported on A, it
can simply return the requestor number (recall that is this is
really a Remote Capability Descriptor, RCD) to A, containing the
fact that the capability specified is one that is local to A and
giving A the index to the capability in its supported capability
list.
C. Security
The mechanism presented in B. brings up something of a security
issue. If B. tries to invoke a capability in A's supported list,
should A allow B access without question? If it did, any host on
the network could maliciously invoke any capability supported by
any other host. To allow access only if it has been granted
through the standard invocation mechanism, each host can maintain
a bit vector indicating which hosts have access to a given
capability. If a host does receive an invalid request, it is an
error condition.
D. Indirection
There is an additional twist on a Loop problem noted in B.. This
variation comes up when A passes a capability to B who then wants
to pass it to C. Here again B may unambiguously specify which
capability is to be passed by simply sending the Remote Capability
Descriptor (RCD) that is knows it by. The RCD indicates that the
capability, however, A would probably not believe that C should
have access to it.
B must tell A. "1, who have access to your 1'th capability, want
to grant it to host C". To do this, another message type is used.
The "Give" message specifies the supported capability and the host
that it should be given to (refer to figure 6). Here again,
giving away a capability that you don't have is an error
condition.
E. Acknowledgement -
There is one last problem with the "Give" message. If B sends the
"Give" message to A and then continues to send the Remote
Capability Descriptor (RCD) to C, C may try to use the RCD before
the "Give" is received by A. For this reason, B must wait until A
has "ACK"nowledged the "Give" message before sending the RCD to C.
This mechanism requires that hosts queue un"ACK"nowledged "Give"s.
The format for an "ACK" is given in figure 6. This queueing may
be avoided for most "Give"s after the first for a given RCD, but
only at the cost of much additional memory and broadcasting
"Delete"s (See F. below).
F. Deletion -
If all the requestors on A for a given capability supported on B
are deleted. A may tell B so that B may:
a. Delete A's validation bit in the bit vector for the specified
capability and
b. If there are no hosts left that require support of the given
capability, the capability may be deleted from the supported
capability list.
This function requires a new "Delete" message.
Figure 6 is a summary of the message formats. Figure7-11 flow chart
the complete DCCS. In the flow charts, abbreviations are used to
indicates the directories:
CSL - Capability Support List
RRL - Remote Request List
IPL - Invocation Process List
The table manipulation is not given in detail. Three tables are
needed. The first is associated with the CSL and contains the bit
vectors indicating access as noted in C. above. The second table is
associated with the RRL. It contains a host number for each active
request. An attempted return on a request by a host other that the
requested host is an error. The final table is a message buffer
containing the pending "Invoke" and "Return" requests.
In order to avoid hazards in referencing the CSL and its table, a
semaphore called the CSLS is used. A message buffer semaphore, MBS,
is similarly used to lock the message buffer. For the RRL and IPL no
locks are needed with the algorithms given.
Generalization and Application
To implement the DCCS, we assumed a network of CCS systems. The
specifications of the CCS were, however, very loose. For example, no
mention was made of instruction sets. Any CCS-like implementation
could use the mechanisms described herein to snare their objects. A
process passed to system with a different instruction set, for
example, could be used as an efficient emulator.
The most important generalization of the DCCS is to note that a given
implementation has no idea what kind of host it is talking to over
the network. Any sort of host could implement a protocol using the
messages given. For example, a single user system might allow its
user to perform arbitrary invocations on remote capabilities and keep
a table of returned capabilities. Such a system might also support
some kind of standard terminal capability that could be given to
remote processes. On a multi-user system, similar functions could be
performed for each user.
In some sense, any system implementing the DCCS protocol becomes a
C-list system. The single user system could, for example, set up
remote processes servicing remote server capabilities giving out
requestors to the single user system or any other systems. Returns
from invocations could appear on the single user's terminal by remote
invocation of the terminal capability, etc..
Implementing the DCCS on non-C-list systems is similar in some
respects to what happened with some host to host protocol
implementations on the Department Of Defense's ARPA network [10].
The ARPA network host to host protocols allows a process on one
system to communicate with a process on another. Many of the ARPA
net protocol implementations had the effect of introducing local
process to process communication in hosts that formerly had none.
Applications
I. Single Copy
The first application is a solution to what I have dubbed the
single copy problem for information resources. Whenever a
process receives information from a information resource, it
can only receive a local copy of the information. This fact is
apparent when the information come from a distributed data
base, but is also true in tightly coupled virtual memory
situations where information from shared memory must be copied
into local registers for processing. Once a process has a
local copy of some information, it might like to try to insure
that the information remains current, i.e., that it is the
single copy.
The traditional solution to this problem is to lock the
information resource with a semaphore before making a local
copy and then invalidate the local copy before unlocking the
resource. This solution suffers from the fact that, even
though other processes may not be requesting the copied data,
the data must be unlocked quickly just in case. This can
result in many needless copies being made.
What is needed is a mechanism for invalidating local copies
exactly when requests by other processes would force
invalidation. To offer such a mechanism, an information
resource can have, in addition to the usual reading and writing
invocations, the following:
"White lock", portion; > ; write notify
"RW lock", portion; >; RW notify
The important invocation on the notify capabilities is:
"Wait for notification"; > reason;
The basic idea is to allow a process to request that it be
notified if an attempt is being made to invalidate its copy.
If the copy is used for reading only, the process need only
request notifications of attempted modifications of the data
("Write lock"). When a process is so notified, it is expected
to invalidate its copy and delete its write notify capability
to inform the information resource server that the pending
write access may proceed.
In the read write lock case, the RW notify capability may also
be used for reading and writing the portion. Any other access
to the portion will cause notification. When notified, the
process with the RW notify capability is expected to write back
the latest copy of the information before deleting its RW
notify capability.
Space does not permit presenting more details for this
mechanism. The important fact to notice is that it permits an
information resource to be shared in such a way that, though
the information may be widely distributed, it is made to appear
as a single copy. This mechanism has important applications to
distributed data bases.
II. Network Resource Optimization
The application that probably best demonstrates the usefulness
of the DCCS is the sort of network optimization capability that
can be used to create at least the primitive capabilities
introduced earlier:
9. Account Capability
a. "Create", type; >; capability
The passed type parameter could at least be any of: "File",
"Directory", "Process", or "Server". The appropriate type
of capability would be returned. The resources used for the
capability are charged to the particular account.
Now suppose that a user on one CCS system within a DCCS network
has remote access to account capabilities on several other CCS
systems. This user could create what might be called a super
account capability to optimize use of his network resources.
The super account capability would actually be a requestor
serviced by a process with optimization desired would be
completely under user control, but some of the more obvious
examples are presented:
1. Static Object Creation Optimization
a. When a new file is requested, create it on the system
with the fastest access or the least cost per bit.
b. When a process is requested, create it on the system with
the fastest current response or with the least cost per
instruction.
2. Dynamic optimization.
To do dynamic optimization, the super account would not give
the requesting process the capability that it received from
the remote account after its static optimization, but would
give out a requestor that it would make function like the
actual capability except optimized.
a. When network conditions or user needs charges, files can
be moved to more effective systems. changes in cost
conditions might result in file movement. Charges in
reliability conditions might result in movement of files
and/or in addition or deletion of multiple copies.
b. If system load conditions or CPU charges change, it might
be effective to relocate a process. The super account
service process could: create a new process on a more
effective system, stop the old process, move the old C-
list and memory to the new process and start the new
process up. The emulation process given to the user
would never appear to change.
c. Similar optimizations can be done on any other
capabilities.
Such a super account can automatically optimize a user's
network resources to suit the user's needs without changing
the functional characteristics of the objects being
optimized.
Final Note
The DCCS mechanisms defined in this paper are currently being
implemented on a Digital Equipment Corporation PDP-11/45 computer for
use as an experimental protocol on the ARPA computer network [10].
The DCCS protocol will also form the basis for a gateway between the
ARPA network and Energy Research and Developement Agency's CTR
network [11]. It is the authors hope that the DCCS mechanism will
hasten the approach of the kind of networks that are needed to create
a truly free market in computational resources.
Acknowledgements
The author would like to thank the administrators and staff of the
Computer Research Project at the Lawrence Livemore Laboratory for
creating the kind of environment conductive to the ideas presented in
this paper. Special thanks are due to Charles Landau for many of the
C-list ideas as implemented in the current RATS system.
References
1. C. R. Landau, The RATS Operating System, Lawrence Livermore
Laboratory, Report UCRL-77378 (1975)
2. C. R. Landau, An Introduction to RATS (RISOS/ARPA Terminal
System): An Operating System for the DEC PDP-11/45, Lawrence
Livermore Laboratory, Report UCRL-51582 (1974)
3. J. E. Donnelley, Notes on RATS and Capability List Operating
Systems, Lawrence Livermore Laboratory, Report UCID-16902 (1975)
4. B. W. Lampson, "On Reliable and Extendable Operating Systems",
Techniques in Software Engineering, NATO Sci Comm. Workshop Material,
Vol. II (1969)
5. W. Wulf, et. al., "HYDRA: The Kernel of a Multiprocessor Operating
System", Communications of the ACM 17 6 (1974)
6. P. Neumann et. al., "On the Design of a Provably Secure Operating
System" International Workshop on Protection in Operating Systems,
IRIA (1974)
7. R. S. Fabry, "Capability-Based Addressing", CACM 17 7 (1974)
8. E. W. Dijkstra, "Cooperating Sequential Processes", published in
Programming Languages, F. Genuys, editor, Academic Press, pp. 43-112
(1968)
9. F. A. Akkoyunlu, et. al., "Some Constraints and Tradeoffs in the
Design of Network Communications", Proceedings of the Fifth Symposium
on Operating System Principles, Vol. 9 No. 5 pp. 67-74 (1975)
10. L. G. Roberts and B. D. Wessler, "Computer Network Development to
Achieve Resource Sharing", AFLPS Conference Proceedings 36, pp.
543-549 (1970)
11. "National CTR Computer Center", Lawrence Livermore Laboratory
Energy and Technology Review, Lawrence Livermore Laboratory UCRL-
52000-75-12, December (1975)
The figures are not included in the online version. Interested
readers can obtain a hardcopy version of the documents including the
figures by requesting a copy of UCRL-77800 from:
Technical Information Department
Lawrence Livermore Laboratory
University of California Livermore, California 94550
Questions or comments would be appreciated and should be directed to
the author:
Though the U.S. mail:
James E. Donnelley
Lawrence Livermore Laboratory L-307
P. O. Box 808
Livermore, California 94550
By telephone:
(415)447-1100 ext. 3406
Via ARPA net mail:
JED@BBN
"This report was prepared as an account of work sponsored by the
United States Government. Neither the United States nor the United
States Energy Research & Development Administration, nor any of their
employees, nor any of their contractors, subcontractors or their
employees, makes any warranty, express or implied, or assumes any
legal liability or responsibility for the accuracy, completeness or
usefulness of any information, apparatus, product or process
disclosed, or represents that its use would not infringe privately-
owned rights."