This page describes the configuration file for the Hercules S/370, ESA/390, and z/Architecture emulator.
The configuration file hercules.cnf contains the processor and device layout. It is roughly equivalent to the IOCDS on a real System/390. The configuration file is an ASCII text file.
Please note that the below example configuration file should not be considered a good example of what an actual configuration file looks like. It is only meant to illustrate what some of the supported configuration file statements look like and how they are used.
|
Blank lines, and lines beginning with a # sign or an asterisk, are treated as comments.
System parameters may appear in any order but they must precede all device statements. Each system parameter must be on a separate line. The following system parameters may be specified:
ARCHMODE S/370 | ESA/390 | z/Arch | ESAME
specifies the initial architecture mode:
S/370 for OS/360, VM/370, and MVS 3.8.
ESA/390 for MVS/XA, MVS/ESA, OS/390, VM/ESA, VSE/ESA,
Linux/390, and ZZSA.
z/Arch or ESAME for z/OS and zLinux.
ESAME is a synonym for z/Arch.
When z/Arch or ESAME is specified,
the machine will always IPL in ESA/390 mode,
but is capable of being switched into z/Architecture mode after IPL.
This is handled automatically by all z/Architecture operating systems.
ASN_AND_LX_REUSE DISABLE | ENABLE
specifies that the ASN-and-LX-Reuse Facility (ALRF) is to be disabled
or enabled. The default is disabled. This is a z/Architecture-only
feature (it is always disabled for S/390 or ESA/390). Set this
to ENABLE if the operating system supports
this z/Architecture feature and the use of this feature is desired.
Set it to DISABLE or do not specify anything
if the operating system doesn't support this feature, and it
inadvertently sets CR0 bit 44 to 1, usually leading to unexpected
program interrupt when instructions such as LASP are issued.
ASN_AND_LX_REUSE may be abbreviated as ALRF
AUTOMOUNT [±]directory
specifies the host system directory where the guest is allowed or not allowed to automatically load virtual tape volumes from. Prefix allowable directories with a '+' plus sign and unallowable directories with a '-' minus sign. The default prefix if neither is specified is the '+' plus sign (i.e. an allowable directory).
Caution: Enabling this feature may have security consequences depending on which allowable host system directories you specify as well as how your guest operating system enforces authorized use of the Set Diagnose (X'4B') channel command code.
All host system virtual tape volumes to be "automounted" by the guest must reside within one of the specified allowable host system directories or any of its subdirectories while not also being within any of the specified unallowable directories or any of their subdirectories, in order for the guest-invoked automount to be accepted.
Note: specifying a disallowed automount directory does not preclude the
Hercules operator from manually mounting any desired file via the
devinit panel command -- even one in a currently defined
"disallowed" automount directory. The AUTOMOUNT statement only controls
guest-invoked automatic tape mounts and not manual tape mounts performed
by the Hercules operator.
All directories must be specified on separate statements, but as many
statements as needed may be specified in order to describe the desired
allowable/unallowable directories layout. For convenience, an
automount panel command is also provided to dynamically
add/remove new/existing allowable/unallowable automount
directories at any time.
The automount feature is activated whenever you specify at least one allowable or unallowable directory. If only unallowable directories are specified, then the current directory becomes the only defined allowable automount directory by default.
All specified directories are always resolved to fully-qualified absolute directory paths before being saved.
Refer to the description of the virtual tape device 'noautomount' option for more information.
AUTO_SCSI_MOUNT NO | YES | nn
specifies whether automatic detection of SCSI tape mounts are to be enabled or not.
Specifying NO or 0 seconds (the default) indicates
the option is disabled, forcing all SCSI tape mounts to be done
manually via an appropriate devinit command.
A value from 1 to 99 seconds inclusive enables the option
and causes periodic queries of the SCSI tape drive to automatically
detect when a new tape is mounted. Specifying YES
is the same as specifying 5 seconds, the current default interval.
The scsimount panel command may also be used to display
and/or modify this value on demand once Hercules has been started. Note
too that the scsimount panel command also lists any mounts
and/or dismounts that may still be pending on the drive, as long as
you've defined your tape drive as a model that has an LCD "display"
(such as a model 3480, 3490 or 3590).
Note: enabling this option may cause Hercules to take longer to shutdown depending on the value specified for this option as well as how the host operating system (Windows, Linux, etc) and associated hardware (SCSI adapter) behaves to drive status queries for drives which do not have any media currently mounted on them.
CCKD cckd-parameters
The CCKD command and initialization statement can be used to affect cckd processing. The CCKD initialization statement is specified as a Hercules configuration file statement and supports the same options as the cckd panel command. Refer to the Compressed Dasd Emulation web page for more information.
CODEPAGE mapping
specifies the codepage conversion mapping table used for ASCII/EBCDIC translation.
default specifies traditional Hercules codepage mapping.
Other supported codepage mappings are:
| Mapping | Description | |
|---|---|---|
| ASCII | EBCDIC | |
437/037 |
437 PC United States | 037 United States/Canada |
437/500 |
437 PC United States | 500 International |
437/1047 |
437 PC United States | 1047 Open Systems Latin 1 |
819/037 |
819 ISO-8859-1 | 037 United States/Canada |
819/037v2 |
819 ISO-8859-1 | 037 United States/Canada version 2 |
819/273 |
819 ISO-8859-1 | 273 Austria/Germany |
819/277 |
819 ISO-8859-1 | 277 Denmark/Norway |
819/278 |
819 ISO-8859-1 | 278 Finland/Sweden |
819/280 |
819 ISO-8859-1 | 280 Italy |
819/284 |
819 ISO-8859-1 | 284 Spain |
819/285 |
819 ISO-8859-1 | 285 United Kingdom |
819/297 |
819 ISO-8859-1 | 297 France |
819/500 |
819 ISO-8859-1 | 500 International |
819/1047 |
819 ISO-8859-1 | 1047 Open Systems Latin 1 |
850/273 |
850 PC Latin 1 | 273 Austria/Germany |
850/1047 |
850 PC Latin 1 | 1047 Open Systems Latin 1 |
1252/037 |
1252 Windows Latin 1 | 037 United States/Canada |
1252/037v2 |
1252 Windows Latin 1 | 037 United States/Canada version 2 |
1252/1047 |
1252 Windows Latin 1 | 1047 Open Systems Latin 1 |
1252/1140 |
1252 Windows Latin 1 | 1140 United States/Canada with Euro |
Iconv single byte codepages may also be used (e.g. UTF8/EBCDIC-CP-NL)
if the host environment supports iconv.
If no codepage is specified then the environment variable HERCULES_CP
will be inspected. The default codepage mapping is default.
CNSLPORT nnnn
specifies the port number (in decimal) to which tn3270 and telnet clients will connect.
The CNSLPORT statement may also have the form of host:port, where the telnet console server will bind to the specified address.
CONKPALV (idle,intv,count)
specifies the tn3270 console and telnet clients keep-alive option values that control automatic detection of disconnected tn3270/telnet client sessions.
idle specifies the number of seconds
of inactivity until the first keep-alive probe is
sent (idle time until first probe, or probe frequency).
intv
specifies the interval in seconds between when successive
keep-alive packets are sent if no acknowledgement is received from
the previous one (i.e. the timeout value of the probes themselves).
count specifies the number of unacknowledged
keep-alive packets sent before the connection is considered to have
failed.
The default values are 3, 1, and 10. That is, send the initial probe 3 seconds after the line goes idle and then wait no more than one second for it to be responded to. Do this 10 times before considering the client as having died.
Note: This is a built-in feature of TCP/IP and allows detection of unresponsive TCP/IP connections and not idle clients. That is to say, your connection will not be terminated after 3 seconds of idle time. Your 3270 session can remain idle for many minutes without any data being transmitted. If the TCP/IP stack at the other end of the connection -- not your 3270 client itself -- fails to respond to the internal keep-alive probe packets, then it means that the TCP/IP stack is down or there has been a break in the connection. Thus, even if your 3270 client is completely idle, your system's TCP/IP stack itself should still respond to the keep-alive probes sent by the TCP/IP stack at the Hercules end of the link. If it doesn't, then TCP/IP will terminate the tn3270/telnet session which will cause Hercules to disconnect the terminal.
The three values can also be modified on-demand via the conkpalv
panel command, which has the exact same syntax. Note that the
syntax is very unforgiving: no spaces are allowed anywhere within the
parentheses and each value must be separated from the other with a
single comma.
Note:
On Windows platforms the count value is ignored and
cannot be changed from its default value of 10.
Also, some older platforms may ignore all of the values specified
and use platform default values instead.
CPUMODEL xxxx
specifies the 4 hexadecimal digit CPU machine type number stored by the STIDP instruction Note: Prior to ESA/390 this was known as the CPU model number
CPUPRIO nn
specifies the priority of the CPU thread. Default is a nice value of 15, which means a low priority such that I/O can be scheduled and completed in favour of CPU cycles being burned. On Multi-CPU systems, a real CPU can be "dedicated" to Hercules, by giving the CPU thread a very high dispatching priority (-20). See "Thread Priorities" below for more information.
Caution: CPUPRIO should not have a higher dispatching priority than the TOD Clock and timer thread.
CPUSERIAL xxxxxx
specifies the 6 hexadecimal digit CPU serial number stored by the STIDP instruction
CPUVERID xx
specifies the 2 hexadecimal digit CPU version code stored by the STIDP instruction. The default version code is FD when ARCHMODE S/370 or ARCHMODE ESA/390 is specified. For the z/Architecture mode, the version code is always stored as 00 and the value specified here is ignored.
DEFSYM symbol value
Defines symbol symbol as to contain value value. The symbol can then be the object of a substitution later in the configuration file or for panel commands. If value contains blanks or spaces, then it should be enclosed in double quotation marks ("). See substitutions for a more in-depth discussion on this feature.
Substitution is available even in configuration statements,
meaning it is possible to perform substitution in the DEFSYM statement itself.
However, symbols are always defined as the last step in the process, so attempting
to self define a symbol will result in an empty string:
DEFSYM FOO $(FOO)
DEVPRIO nn
specifies the priority of the device threads. The default value is 8. See "Thread Priorities" below for more information.
Caution: DEVPRIO should not have a higher dispatching priority than the TOD Clock and timer thread.
DEVTMAX -1 | 0 | nnn
specifies the maximum number of device threads allowed.
Specify -1 to cause 'one time only' temporary threads to be
created to service each I/O request to a device. Once the I/O request is
complete, the thread exits. Subsequent I/O to the same device will cause
another worker thread to be created again.
Specify 0 to cause an unlimited number of 'semi-permanent'
threads to be created on an 'as-needed' basis. With this option, a thread
is created to service an I/O request for a device if one doesn't already
exist, but once the I/O is complete, the thread enters an idle state waiting
for new work. If a new I/O request for the device arrives before the timeout
period expires, the existing thread will be reused. The timeout value is
currently hard coded at 5 minutes. Note that this option can cause one thread
(or possibly more) to be created for each device defined in your
configuration. Specifying 0 means there is no limit to the
number of threads that can be created.
Specify a value from 1 to nnn to set an upper limit
to the number of threads that can be created to service any I/O request to
any device. Like the 0 option, each thread, once done servicing
an I/O request, enters an idle state. If a new request arrives before the
timeout period expires, the thread is reused. If all threads are busy when
a new I/O request arrives however, a new thread is created only if the
specified maximum has not yet been reached. If the specified maximum number
of threads has already been reached, then the I/O request is placed in a queue
and will be serviced by the first available thread (i.e. by whichever thread
becomes idle first). This option was created to address a threading issue
(possibly related to the cygwin Pthreads implementation) on Windows systems.
The default for Windows is 8. The default for all other systems
is 0.
DIAG8CMD DISABLE | ENABLE [ECHO | NOECHO]
When ENABLE is specified, commands issued through the Diagnose 8 interface
will be executed by Hercules as Hercules commands. When set to DISABLE,
commands issued through the Diagnose 8 interface will generate a Specification
Exception program interrupt on the issuing CPU.
An optional second argument can be given to request whether the commands
issued using the Diagnose 8 interface will be traced at the console. This may be
useful for programs that routinely issue panel commands using the Diagnose 8 interface.
When ECHO is specified, a message is issued as the panel is about to issue
the command, the command is redisplayed as if it was entered through the panel input
line, and a final message is issued to indicate the command completed. When NOECHO
is specified, no such messages are displayed and the command completes silently.
The value of ECHO or NOECHO has no effect on
command output being placed into a response buffer if the Diagnose 8 interface
requested one.
The default is DISABLE NOECHO
Caution: Enabling this feature may have security consequences.
When this feature is enabled it is possible for guest operating systems
running under Hercules to issue commands directly to the host operating system
by means of the Hercules sh (shell) command. This ability may be
disabled via the SHCMDOPT statement.
ECPSVM YES | NO | LEVEL nn
specifies whether ECPS:VM (Extended Control Program Support : Virtual Machine)
support is to be enabled. If YES is specified, then the support
level reported to the operating system is 20. The purpose of
ECPS:VM is to provide to the VM/370 Operating system a set of shortcut
facilities to perform hypervisor functions (CP Assists) and virtual
machine simulation (VM Assists). Although this feature does not affect
VM Operating system products operating in XA, ESA or z/Architecture mode,
it will affect VM/370 and VM/SP products running under VM/XA, VM/ESA or z/VM.
Running VM/370 and VM/SP products under VM/XA, VM/ESA or z/VM should be
done with ECPS:VM disabled. ECPS:VM should not be enabled in an AP or MP
environment. ECPS:VM has no effect on non-VM operating systems. It is
however recommended to disable ECPS:VM when running native non-VM operating
systems. If a specific LEVEL is specified, this value will be reported
to the operating system when it issues a Store ECPS:VM level, but it
doesn't otherwise alter the ECPS:VM facility operations. This is a partial
implementation.
ENGINES [nn*]CP|IL|AP|IP[,...]
specifies the type of engine for each installed processor. The default engine type is CP.
nn* is an optional repeat count. Spaces are not permitted.
Examples:
ENGINES CP,CP,AP,IP
specifies that processor engines 0 and 1 are of type CP, engine 2 is
type AP, and engine 3 is type IP.
ENGINES 4*CP,2*AP,2*IP
specifies that the first four processor engines (engines 0-3) are of
type CP, the next two (engines 4-5) are of type AP, and the next two
(engines 6-7) are of type IP.
The number of installed processor engines is determined by the MAXCPU statement. If the ENGINES statement specifies more than MAXCPU engines, the excess engines are ignored. If fewer than MAXCPU engines are specified, the remaining engines are set to type CP.
HERCPRIO nn
specifies the process priority for Hercules. The default is 0. See "Process Priorities" below for more information.
HTTPPORT nnnn [AUTH | NOAUTH] [ userid password ]
specifies the port number (in decimal) on which the HTTP server
will listen. The port number must either be 80
or within the range 1024 - 65535 inclusive. If no HTTPPORT statement is
present or an invalid port number is specified, then the HTTP server thread
will not be activated.
AUTH indictates that a userid and password are required to access
the HTTP server, whereas NOAUTH indicates that a userid and password
are not required. The userid and password may be any valid string.
HTTPROOT directory
specifies the root directory where the HTTP server's files reside. If not specified, the default value for Win32 builds of Hercules is the directory where the Hercules executable itself is executing out of, and for non-Win32 builds it is the directory specified as the default package installation directory when the Hercules executable was built (which can vary depending on how the Hercules package was built, but is usually /usr/local/share/hercules/).
IGNORE INCLUDE_ERRORS
Indicates that errors caused by subsequent
INCLUDE statements
for files which do not exist should instead be ignored rather
than causing startup to be aborted (as would otherwise normally
occur).
INCLUDE filepath
An INCLUDE statement tells Hercules configuration file
processing to treat the contents of the file specified by filepath
as if its contents had appeared in the configuration file at the point
where the INCLUDE statement appears.
Note that the included file may itself contain yet another
INCLUDE statement as long as the maximum nesting depth
(current 8) is not exceeded.
IODELAY usec [NOWARN]
specifies the amount of time (in microseconds) to wait after
an I/O interrupt is ready to be set pending. This value can also be
set using the Hercules console. The purpose of this parameter is to
bypass a bug in the Linux/390 and zLinux dasd.c
device driver. The problem is more apt to happen under Hercules than
on a real machine because we may present an I/O interrupt sooner than a
real machine.
If the IODELAY value is non-zero a warning message (HHCCF037W) will be
issued unless NOWARN is specified.
NOTE : OSTAILOR LINUX no longer sets
IODELAY to 800 since the problem described above is no longer present in
recent versions of the Linux kernel.
LDMOD module list
specifies additional modules that are to be loaded by the Hercules dynamic loader. The default search order is with the hercules directory in the default DLL search path. Most systems also support absolute filenames (ie names starting with '/' or '.') in which case the default search path is not taken.
Multiple LDMOD statements may be used.
LEGACYSENSEID OFF | DISABLE | ON | ENABLE
specifies whether the SENSE ID CCW (X'E4') will be honored for the devices that originally didn't support that feature. This includes (but may not be limited to) 3410 and 3420 tape drives, 2311 and 2314 direct access storage devices, and 2703 communication controllers.
Specify ON or ENABLE if your guest
operating system needs the Sense ID support to dynamically
detect those devices. Note that most current operating systems
will not detect those devices even though Sense ID is enabled
because those devices never supported the Sense ID in the first
place. So this mainly applies to custom built or modified versions
of guest operating systems that are aware of this specific Hercules
capability.
Because those legacy devices didn't originally support this command,
and for compatibility reasons, the default is OFF
or DISABLE.
LOADPARM xxxxxxxx
specifies the eight-character IPL parameter which is used by some operating systems to select system parameters.
LOGOPT TIMESTAMP | NOTIMESTAMP
sets Hercules log options. TIMESTAMP causes messages to the log to be time stamped. NOTIMESTAMP prevents time stamping of log messages. TIMESTAMP and NOTIMESTAMP may be abbreviated as TIME and NOTIME respectively. The current resolution of the stamp is one second.
The default is TIMESTAMP.
LPARNAME name
specifies the LPAR name returned by DIAG X'204'. The default is
HERCULES.
LPARNUM xx
specifies the one- or two-digit hexadecimal LPAR identification number stored by the STIDP instruction. If a one-digit number is specified then STIDP stores a format-0 CPU ID. If a two-digit number is specified then STIDP stores a format-1 CPU ID. If LPARNUM is not specified, then STIDP stores a basic-mode CPU ID.
MAINSIZE nnnn
specifies the main storage size in megabytes, where
nnnn is a decimal number. The lower limit is 2.
The actual upper limit is determined by your host system's architecture
and operating system, and (on some systems) the amount of physical
memory and paging space you have available.
MANUFACTURER name
specifies the MANUFACTURER name returned the STSI instruction. The default is
HRC.
MAXCPU nn
specifies the number of installed processor engines. The NUMCPU statement specifies the number of engines which will be configured online at startup time. All processors are CP engines unless otherwise specified by the ENGINES statement.
The value of MAXCPU cannot exceed the value of MAX_CPU_ENGINES.
If MAXCPU is not specified then the default value is the value of
MAX_CPU_ENGINES.
MAX_CPU_ENGINES is a compile-time variable which sets
an upper limit on the value of MAXCPU.
The value of MAX_CPU_ENGINES is displayed in the
Build information message on the Hercules control panel at startup time.
To change the value of MAX_CPU_ENGINES you must rebuild
Hercules.
For Unix builds, specify
./configure --enable-multi-cpu=nn
before performing make.
For Windows builds, specify
SET MAX_CPU_ENGINES=nn
before performing nmake.
MAX_CPU_ENGINES may be up to 128 on 64-bit Linux platforms.
On Windows, and on all 32-bit platforms, the maximum value is 64.
For performance reasons,
values above 32 are not recommended for 32-bit platforms.
If MAX_CPU_ENGINES is set to 1 then multiprocessing is disabled.
See also NUMCPU for a discussion of the performance
implications of MAX_CPU_ENGINES.
MODEL hardware_model
[ capacity_model ]
[ perm_capacity_model ]
[ temp_capacity_model ]
specifies the MODEL names returned by the STSI instruction.
If two operands are supplied, the first is the hardware model name (CPC ND model) and the second is the capacity model name (CPC SI model). If only one operand is supplied, it is used as both the hardware model name and the capacity model name. The optional third and fourth operands specify the permanent capacity model name and the temporary capacity model name respectively.
The default is EMULATOR.
MODPATH path
specifies the path where dynamic modules are loaded from. When a modpath statement is specified, the path on the modpath statement is searched before the default path is searched. When a relative path is specified is interpreted as a relative path within the default search path, if an absolute path is specified is interpreted as such.
The default MODPATH is hercules, which means modules are loaded from the directory hercules within the default LD_LIBRARY_PATH.
MOUNTED_TAPE_REINIT DISALLOW | ALLOW
specifies whether reinitialization of tape drive devices (via the
devinit command, in order to mount a new tape) should
be allowed if there is already a tape mounted on the drive.
Specifying ALLOW (the default) indicates new tapes may
be mounted (via 'devinit nnnn new-tape-filename')
irrespective of whether or not there is already a tape mounted on the drive.
Specifying DISALLOW prevents new tapes from being mounted
if one is already mounted. When DISALLOW is specified
and a tape is already mounted on the drive, it must first be unmounted
(via the command 'devinit nnnn *') before
the new tape can be mounted. Otherwise the devinit attempt to mount
the new tape is rejected.
This option is meant as a safety mechanism to protect against accidentally dismounting a tape from the wrong drive as a result of a simple typo (thereby cancelling a potentially important tape job) and was added by user request.
Also note that for SCSI tape drives the 'devinit nnnn *'
command has no effect as the tape must be unmounted manually (since it is
a real physical device and not one emulated via a disk file like .AWS tapes).
NUMCPU nn
specifies the number of emulated processor engines which will be configured online at startup time. NUMCPU cannot exceed the value of MAXCPU. If NUMCPU is less than MAXCPU then the remaining engines can be configured online later.
Multiprocessor emulation works best if your host system actually has more than one physical CPU, but you can still emulate multiple CPUs nervertheless even on a uniprocessor system (and you might even achieve a small performance benefit when you do). There is little point, however, in specifying NUMCPU greater than 1 unless your guest operating system (running under Hercules) is actually able to support multiple CPUs (and if you do not actually need multiprocessor emulation, then setting MAX_CPU_ENGINES to 1 at compile time might even produce a slight performance advantage too).
NUMVEC nn
specifies the number of emulated vector facilities. Default is one per CPU. Only available by default in ESA/390 mode.
OSTAILOR OS/390 | z/OS |
VM | VSE | LINUX | QUIET | NULL
specifies the intended operating system. The effect of this
parameter is to reduce control panel message traffic by
selectively suppressing trace messages for program checks
which are considered normal in the specified environment.
QUIET discards all exception messages.
NULL allows all exception messages to be logged.
Optionally prefix any value except QUIET or
NULL with '+' to cause the suppressions for that
environment to be combined (added) to those already specified,
or with '-' to remove such suppressions (i.e. to allow them).
If the OSTAILOR
statement is omitted, exception messages for program checks
10, 11, 16, and 1C are suppressed.
Use the ostailor or pgmtrace panel
commands to display or alter the current settings.
PANRATE SLOW | FAST | nn
specifies the panel refresh rate, in milliseconds between refreshes. SLOW is the same as 500, and FAST is the same as 50. A value less than the Linux system clock tick interval (10 on Intel, 1 on Alpha), or more than 5000, will be rejected. SLOW is the default.
PANTITLE "title-string"
specifies an optional console window title-bar string to be used in place of the default supplied by the windowing system. If the value contains any blanks it must be enclosed within double-quotes.
This option allows one to distinguish between different Hercules sessions when running more than one instance of Hercules on the same machine.
This option takes effect only when the Hercules console is displayed on an xterm terminal (commonly used on Unix systems), or in a Windows command prompt window. Note that this option has no effect when Hercules is run under control of the Hercules GUI since Hercules's console window is hidden in favor of using the GUI's window instead.
PGMPRDOS RESTRICTED | LICENSED
specifies whether or not Hercules will run licensed program product ESA
or z/Architecture operating systems. If RESTRICTED is
specified, Hercules will stop all CPUs when a licensed program product
operating system is detected. Specify
LICENSED to allow these operating systems to run normally.
This parameter has no effect on Linux/390, Linux for z/Series, or any
370-mode OS.
NOTE: It is YOUR responsibility to comply with the terms of the license for the operating system you intend to run on Hercules. If you specify LICENSED and run a licensed operating system in violation of that license, then don't come after the Hercules developers when the vendor sends his lawyers after you.
RESTRICTED is the default. Specifying
LICENSED will produce a message when a licensed operating
system is detected to remind you of your responsibility to comply with
software license terms.
PLANT name
specifies the PLANT name returned by the STSI instruction. The default is
ZZ.
SHCMDOPT DISABLE | NODIAG8
When set to DISABLE, sh (shell) commands are globally disabled, and will result
in an error if entered either directly via the Hercules hardware console or
programmatically via the DIAG8CMD interface.
When set to NODIAG8 only the programmatic execution of shell commands via the
the Diagnose 8 interface is disabled, but sh (shell) commands entered directly
via the Hercules hardware console will still work.
NOTE: "entered directly via the Hercules hardware console" also pertains to both commands entered via the HTTP server facility as well as commands entered via .rc "run command" scripts.
SHRDPORT nnnn
specifies the port number (in decimal) on which the Shared Device server
will listen. Specifying SHRDPORT will allow other Hercules instances
to access devices on this instance. (Currently only DASD devices may
be shared). By default, the other Hercules instances (clients) will
use port 3990. If you specify a different port number, then you will
have to specify this port number on the device statement for the other
Hercules clients.
If no SHRDPORT statement is present then the Shared Device server thread
will not be activated.
SYSEPOCH yyyy [±years]
specifies the base date for the TOD clock. Use the default value
(1900) for all systems except OS/360. Use 1960
for OS/360. Values other than these were formerly used to offset the
TOD clock by a number of years to move the date before the year 2000 for
non-Y2K-compliant operating systems. This use is deprecated, and support
will be removed in a future release; at that time, only values of
1900 or 1960 will be accepted. Other values
will produce a warning message with the equivalent values to specify in
the SYSEPOCH statement.
An optional year offset may be specified, and will be treated as though
it had been specified on a YROFFSET
statement.
TIMERINT DEFAULT | nnnn
specifies the internal timers update interval, in microseconds. This parameter specifies how frequently Hercules's internal timers-update thread updates the TOD Clock, CPU Timer, and other architectural related clock/timer values. The default interval is 50 microseconds, which strikes a reasonable balance between clock accuracy and overall host performance. The minimum allowed value is 1 microsecond and the maximum is 1000000 microseconds (i.e. one second).
Caution: While a lower TIMERINT value may help increase the accuracy of your guest's TOD Clock and CPU Timer values, it could also have a severe negative impact on the overall performance of your host operating system. This is especially true when a low TIMERINT value is coupled with a high HERCPRIO and TODPRIO priority setting. Exercise extreme caution when choosing your desired TIMERINT in relationship to your chosen HERCPRIO and TODPRIO priority settings.
TODDRAG nn
specifies the TOD clock drag factor. This parameter can be used to slow down or speed up the TOD clock by a factor of nn. A significant slowdown can improve the performance of some operating systems which consume significant amounts of CPU time processing timer interrupts. A drag factor of 2.0 slows down the clock by 50%. A drag factor of 0.5 doubles the speed of the clock. A drag factor of 1.01 slows down the clock by 1%, and 0.99 speeds up the clock by 1%.
TODPRIO nn
specifies the priority of the TOD Clock and timer thread. The default value is -20. See "Thread Priorities" below for more information.
Caution: TODPRIO should be given a dispatching priority equal to or higher than any other thread within Hercules.
TRACEOPT TRADITIONAL | REGSFIRST | NOREGS
sets the Hercules instruction tracing display option.
TRADITIONAL (the default), displays the registers following
the instruction about to be executed such that pressing enter (to execute
the displayed instruction) then shows the next instruction to be executed
followed by the updated registers display.
REGSFIRST displays the current register contents followed by
the instruction about to be executed such that pressing enter (to execute
the displayed instruction) then shows the updated registers followed by
the next instruction to be executed.
NOREGS suppresses the registers display altogether
and shows just the instruction to be executed.
In addition to the TRACEOPT configuration file statement
there is also a corresponding traceopt panel command to
dynamically display and/or update the current setting at any time.
TZOFFSET ±hhmm
specifies the hours and minutes by which the TOD clock will
be offset from the current system time. For GMT, use the
default value (+0000). For timezones west of Greenwich, specify
a negative value (example: -0500 for US Eastern Standard
Time, -0800 for US Pacific Standard Time).
For timezones east of Greenwich, specify a positive value
(example: +0100 for Central European Time,
+0930 for South Australian Time).
XPNDSIZE nnnn
specifies the expanded storage size in megabytes, where
nnnn is a decimal number. The lower limit is 0.
The actual upper limit is determined by your host system's architecture
and operating system, and (on some systems) the amount of physical
memory and paging space you have available.
YROFFSET ±years
specifies a number of years to offset the TOD clock from the actual
date. Positive numbers will move the clock forward in time, while
negative numbers will move it backward. A common value for
non-Y2K-compliant operating systems is YROFFSET -28, which
has the advantage that the day of the week and the presence or absence
of February 29 is the same as the current year. This value may not be
specified as greater than ±142 years, the total range of the TOD
clock. Specifying a value that causes the computed TOD clock year to be
earlier than the value of SYSEPOCH
or more than 142 years later than that value will produce unexpected
results.
A comment preceded by a # sign may be appended to any system parameter statement.
In configuration and device statements, as well as in panel commands and OAT files, symbols may be substituted for text.
To substitute symbol symbol with its contents, the symbol should be enclosed within parenthesis and preceded by a $ sign. For example, if symbol FOO contains the text string "BAR" then $(FOO) will be substituted with the string "BAR";. Symbol names are case sensitive.
DEFSYM TAPEDIR "/home/hercules/tapes"
...
0380 3420 $(TAPEDIR)/scratch.aws
...In this example, device 0380 will be a 3420 loaded with the AWS tape file in /home/hercules/tapes/scratch.aws
When multiple devices are defined with a single device definition statement, then the symbols
| (3 digits device number, upper case hexadecimal digits) | |
| (4 digits device number, upper case hexadecimal digits) | |
| (3 digits device number, lower case hexadecimal digits) | |
| (4 digits device number, lower case hexadecimal digits) |
are defined to contain for each device the relevant device address. For example:
0200,0201 3340 /home/hercules/dasds/myvols.$(CUU)
will define two 3340 packs, with device 0200 being loaded with the file myvols.200 and device 0201 defined with myvols.201.
If a symbol is not explicitly defined by a DEFSYM statement and an environment variable by the same name exists, the string contents of that environment variable will be used for substitution.
If a symbol is not defined by an explicit DEFSYM, is not an automatically generated symbol and is not an environment variable, an empty string will be substituted.
To be able to specify the '$(' string without incurring substitution, an additional '$' sign should be used. For example, $$(FOO) will not be substituted. If substitution is required but the preceding text is to contain a '$' sign as the very last character, then $$$(FOO) should be specified. Thus, if symbol FOO contains "BAR", then $$(FOO) will remain "$$(FOO)" while $$$(FOO) will become "$BAR".
Substitution is not recursive (only one substitution pass is made).
Enhanced symbol substitution differs from the above normal symbol substitution in several very important ways:
First, the syntax is different. Enhanced substitution symbol names are specified
using ${var} (dollar + brace) rather than $(var)
(dollar + parenthesis).
Second, the enhanced syntax supports specifying a default value that is to be used
instead whenever the name symbol is otherwise not defined. The default value is
placed within the opening and closing braces just as the symbol name is, but
separated from it by either a single equal sign '=' or a
colon-equal-sign ':='.
For example, specifying "${DASD_PATH=dasd/}" in your configuration
file requests that the value of the "DASD_PATH" symbol or environment variable be
substituted, or, if the variable is undefined, to use the value "dasd/"
instead. If no default value is specified then an empty string is used instead.
Finally, enhanced symbol substitution occurs only from host defined environment
variables and not from any identically named DEFSYM symbol
should one exist. For example, if environment variable 'FOO' is defined with the
value "bar", then the configuration file statement "DEFSYM FOO myfoo"
followed immediately by the statement "${FOO}" causes the value
"bar" to be substituted and not 'myfoo' as might
otherwise be believed, whereas the statement "$(FOO)", since it is
a normal symbol substitution sequence does get replaced with "myfoo"
(since that was the value defined to it via the preceding DEFSYM
statement).
In other words each symbol substitution technique is supported completely separately
from one another. DEFSYM allows one to define/undefine/use private (internally
defined) symbols separate from the host operating system's environment variable pool,
whereas the enhanced symbol substitution does not and instead only allows read-only
access to the host's environment variable pool with no support for modifying an already
defined symbol (environment variable) but a nonethless convenient means of defining
a default value to be used should the specified host environment variable be currently
undefined.
Further note that symbol names, being the names of environment variables, are subject
to whatever case sensitivity or case insensitivity that the host operating system
happens to enforce/allow. On Windows, environment variables are not case sensitive, whereas on other
operating systems they may be. Thus "${FOO}", "${foo}",
"${Foo}", etc, all cause the same value to be substituted on Windows,
whereas the DEFSYM symbols $(FOO) and $(foo),
being two completely different and unique symbols, could be substituted with two
completely different values (since DEFSYM is
case sensitive across all supported platforms, including Windows).
To substitute symbol symbol with the current environment variable value,
the symbol should be enclosed within braces and preceded by a $ sign. For example,
if an environment variable named FOO holds the value "BAR", then
${FOO} will be substituted with the string "BAR". If the environment variable
"FOO" is not defined then a null (empty) string is substituted instead.
If the string "${FOO:=myfoo}" is used instead, then the value "BAR" will still be
substituted if the value "BAR" was indeed previously assigned to FOO, but will be
substituted with the value "myfoo" instead if the environment variable
FOO is currently undefined.
Note too that the default value is a literal
string and no substitution is applied to it. Thus attempting to use the syntax
"${foo=${bar}}" will not yield the expected results. It will
not be substituted with the currently defined value of the "bar" environment
variable, but rather will always be substituted with the literal string
"${bar" followed immediately by the literal character '}'.
Symbol names (environment variable names) are not case sensitive on Windows whereas they might be on other host operating systems.
Note: Under Linux, a process is a thread and thread priority information applies instead.
For Windows, the following conversions are used for translating Unix process priorities to Windows process priority classes:
| Unix Priority |
Windows Process Priority Class |
Meaning |
|
|---|---|---|---|
| -20 to -16 | Real-time | Process that has the highest possible priority. The threads of the process preempt the threads of all other processes, including operating system processes performing important tasks. For example, a real-time process that executes for more than a very brief interval can cause disk caches not to flush or cause the mouse to be unresponsive. | |
| -15 to -9 | High | Process that performs time-critical tasks that must be executed immediately. The threads of the process preempt the threads of normal or idle priority class processes. An example is the Task List, which must respond quickly when called by the user, regardless of the load on the operating system. Use extreme care when using the high-priority class, because a high-priority class application can use nearly all available CPU time. | |
| -8 to -1 | Above Normal | Process that has priority above the Normal class but below the High class. | |
| 0 to 7 | Normal | Process with no special scheduling needs. | |
| 8 to 14 | Below Normal | Process that has priority above the Idle class but below the Normal class. | |
| 15 to 20 | Idle | Process whose threads run only when the system is idle. The threads of the process are preempted by the threads of any process running in a higher priority class. An example is a screen saver. The idle-priority class is inherited by child processes. |
Caution: On Windows, the value you choose for your Process Priority has a direct impact on how your Thread Priorities are interpreted! You should never modify one without understanding what impact your doing so might have on the other!
On a Linux/Unix host, Hercules needs to be a setuid root
program to allow it to reset its dispatching priority to a high
(negative) value
(i.e., chown root.root hercules; chmod +s hercules).
For Windows, the following conversions are used for translating Linux/Unix thread priorities to Windows thread priorities:
| Unix Priority |
Windows Thread Priority |
Meaning |
|
|---|---|---|---|
| -20 to -16 | Time Critical | Base priority of 15 for Idle, Below Normal, Normal, Above Normal, or High class processes, and a base priority of 31 for Realtime class processes. | |
| -15 to -9 | Highest | Priority 2 points above the priority class. | |
| -8 to -1 | Above Normal | Priority 1 point above the priority class. | |
| 0 to 7 | Normal | Normal priority for the priority class. | |
| 8 to 14 | Below Normal | Priority 1 point below the priority class. | |
| 15 to 19 | Lowest | Priority 2 points below the priority class. | |
| 20 | Idle | Base priority of 1 for Idle, Below Normal, Normal, Above Normal, or High class processes, and a base priority of 16 for Realtime class processes. |
Caution: On Windows, your Thread Priority is interpreted differently based on your chosen Process Priority setting! You should never modify your Thread Priority settings without first reviewing your chosen Process Priority setting!
The remaining statements in the configuration file are device statements. There must be one device statement for each I/O device or group of identical I/O devices. The format of the device statement is:
devnum(s) devtype [ arguments ] [ # comments... ]
where the generic syntax for device numbers is
[n:]CCUU[,CCUU][-CCUU][.nn][...]
as explained below:
devnum(s)
is either a single devnum, a range of devnums (separated by a '-' (dash)), a count of devnums (separated by a '.' (dot/period/stop)), or a comma separated list of devnums. Examples would be 200-210 or 0300.10 or 0400,0403 or 0100,0110-011F.
All devices defined when devnums specifies more than one device have identical characteristics (except for the device number itself). All devices defined as a group must be defined on a single channel. A channel is defined as a contiguous group of 256 (or hexadecimal 100) devices. 0010 and 0020 are on the same channels. 0100 and 0210 are not.
See devnum immediately below for an explanation of how each device number is specified.
The 4 special subtitution symbols CUU, CCUU, cuu and ccuu are also defined for each device in a device group. See substitutions for details.
devnum
is either a 1 to 4 digit hexadecimal number in the range 0000 to FFFF for ESA/390, or 0000 to 0FFF for S/370. The device number uniquely identifies each device to the operating system.
Channel Set / Logical Channel Subsystem
An optional Channel Set or Logical Channel Subsystem Identification can be specified for a device number or group of devices. The Identification number is specified at the beginning of the definition, followed by a ':' character. For example :
1:0400-040F 3270
defines 3270 devices 400 to 40F to be on S/370 Channel Set 1 or on S/390 or z/Architecture Logical Channel Subsystem # 1.
Since each Logical Channel Subsystem defines its own device numbering space, care should be taken in S/370 mode as to define a coherent set of device numbers.
Not all S/390 or z/Architecture operating systems support Multiple Logical Channel Subsystems (this feature was introduced with the z9-109).
If no Channel Set or Logical Channel Subsystem Identification is specified, then it is assumed to be 0.
devtype
is the device type. Valid device types are shown in the table just below.
arguments
is a list of parameters whose meaning depends on the device type. The arguments required for each class of device are shown further below.
# comments...
A comment preceded by a # sign may be appended to any device definition statement.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||
There are no required arguments for this particular device type, but there are however several optional arguments which are discussed below.
To use this device, a tn3270 client must connect to the host machine via the port number specified on the CNSLPORT statement. A valid tn3270 device type, such as IBM-3278, must be used.
If your tn3270 client software allows you to specify a device type suffix
(e.g. IBM-3278@001F ), then you can use the suffix to connect
to that specific device number, if eligible. If no suffix is specified,
then your client will be connected to the first available 3270 device for
which it is eligible, if any.
If you specify a specific terminal device address (via the device type suffix of your tn3270 client software), then you must be eligible to connect at that device address or your connection is immediately rejected; an alternative terminal device for which you might be eligible is not automatically selected instead.
Optional arguments:
groupname
If a terminal group name is given on the device statement, a device type
suffix with this group name can be used to indicate that a device in this
group is to be used. If a group name is specified as a terminal type suffix
(e.g. IBM-3278@GROUPNAME ) and there are no devices defined
for that group (or there are no more available devices remaining in that
group), then the connection is rejected. If no group name is specified
as a terminal type suffix, then the connection will only be eligible for
any terminal devices which do not have a group name specified on
their device statements. The terminal group name, if specified, should
be 1 to 8 alphanumeric characters, the first character being alphabetic,
and it should not be a hexadecimal number. Upper and lower case
letters in the group name are considered to be equivalent.
ipaddr [ mask ]
The optional IP address and optional subnet mask specify the ip address(es) of which client(s) are allowed to connect at the device address identified by the device statement on which they appear. This provides an alternative and/or additional means of specifying to which device(s) a client tn3270 session may, or should, connect.
If the IP address of the tn3270 client trying to connect, when 'and'ed with the optional subnet mask (which defaults to 255.255.255.255 if not specified), matches the IP address entered on the device statement, then the client is eligible to connect at that device address. Otherwise the client is ineligible to connect at that address and then next available device, if any, for which the client is eligible to connect (if any) is selected instead.
If no permissible terminal devices remain (i.e. terminal devices for which the client is eligible to connect), or there are no more available terminal devices remaining, then the client connection is rejected.
The optional IP address and subnet mask may also be specified in conjunction with the previously mentioned terminal group argument, but the terminal group argument, if specified, must be specified ahead of (i.e. before) the optional ip address and subnet mask arguments. To specify an IP address and subnet mask without also specifying a terminal group, simply use '*' as the group name instead.
If an IP address / subnet mask are not specified, then any client tn3270 session is allowed to connect to the device (provided they are also a member of the specified terminal group, if any).
The terminal group name argument, if specified, always takes precedence over any optional ip address and subnet mask which may also be specified.
The integrated 3270 (SYSG) console is similar to a local non-SNA 3270 device, except that it is not addressed by subchannel number and it is supported only by certain system control programs. The SYSG console is defined like a 3270 device except that the device type is SYSG and the device number is ignored. Only one SYSG console can be defined in a configuration.
Use tn3270 client software to connect to the SYSG console device via the port number specified on the CNSLPORT statement, just as you would connect to a regular local non-SNA 3270 device.
The SYSG console configuration statement recognizes optional arguments which specify group name and IP address in the same way as previously described for a local non-SNA 3270 device. These optional arguments provide a means to ensure that a given tn3270 client can connect directly to the SYSG console. If the group name and IP address arguments are not specified, then the SYSG console is considered to be a member of the general pool of devices eligible for connection to any incoming tn3270 client.
There is one optional argument which is the command prefix for sending input to the device. The default command prefix is '/'.
Note: There is no restriction on the character you can select. If you select a command character that is the first character of a panel command, you will not be able to use that command.
To send a logon command to a 1052-C or 3215-C enter /logon on the Hercules console.
All integrated devices must use a different command prefix.
There are no required arguments for this particular device type, but there are however several optional arguments discussed below.
To use this device, a telnet client must connect to the host machine via the port number specified on the CNSLPORT statement.
If your telnet client software allows you to specify a device type suffix
(for example: ansi@0009 ), then you can use that suffix to specify
the specific 1052 or 3215 device to which you wish to connect. If you do not
specify a suffix in your telnet client software (or your software does not
allow it), then your client will be connected to the first available 1052 or
3215 device for which it is eligible.
An optional noprompt argument may be specified on the device
statement to cause suppression of the "Enter input for console device nnnn"
prompt message which is otherwise normally issued to the device whenever
the system is awaiting input on that device.
Additionally, a terminal group name, ip address and subnet mask may all also be optionally specified in the exact same manner as discussed in the previous Local non-SNA 3270 devices section with the exception that the "noprompt" option, if specified, must precede the other arguments.
The argument specifies a list of file names containing card images. Additional arguments may be specified after the file names:
sockdev
indicates the card reader is a socket device wherein the
filename is actually a socket specification instead of a
device filename. When used, there must only be one filename
specified in the form: port or host:port
or sockpath/sockname. The device then accepts
remote connections on the given TCP/IP port or Unix Domain
Socket, and reads data from the socket instead of from a device
file. This allows automatic remote submission of card reader
data. See the Hercules Socket Reader
page for more details.
eof
specifies that unit exception status is presented after
reading the last card in the file. This option is persistent, and
will remain in effect until the reader is reinitialized with the
intrq option.
intrq
specifies that unit check status with intervention required
sense bytes is presented after reading the last card
in the file. This option is persistent, and will remain in
effect until the reader is reinitialized with the eof
option.
multifile
specifies, when multiple input files are entered, to automatically open the next input file and continue reading whenever EOF is encountered on a given file. If not specified, then reading stops once EOF is reached on a given file and an attention interrupt is then required to open and begin reading the next file.
ebcdic
specifies that the file contains fixed length 80-byte EBCDIC records with no line-end delimiters.
ascii
specifies that the file contains variable length lines of ASCII characters delimited by LF (line feed) sequences or CRLF (carraige return line feed) sequences at the end of each line.
If neither EBCDIC nor ASCII is specified, then the device handler attempts to detect the format of the card image file when the device is first accessed. Auto-detection is not supported for socket devices, and the default is ASCII if sockdev is specified.
trunc
specifies, for ASCII files, that lines longer than 80 characters are truncated instead of producing a unit check error.
autopad
specifies, for EBCDIC files, that the file is automatically padded to a multiple of 80 bytes if necessary.
The argument specifies the name of a file to which the punched output will be written. Additional arguments may be specified after the file name:
ascii
specifies that the file will be written as variable length
lines of ASCII characters delimited by line feeds or
carriage return line feed sequences at the end of each line.
Trailing blanks are removed from each line.
If the ascii argument is not specified, the
file is written as fixed length 80-byte EBCDIC records with
no line-end delimiters.
crlf
specifies, for ASCII files, that carriage return line feed
sequences are written at the end of each line.
If the crlf argument is not specified, then
line-feeds only are written at the end of each line.
noclear
specifies that the output file will not be cleared to zero
bytes when it is opened.
If the noclear argument is not specified, then
any previous content of the file is destroyed when the file
is opened for output.
The argument specifies the name of a file to which the printer output will be written. The output is written in the form of variable length lines of ASCII characters delimited by line feeds or by carriage return line feed sequences. Trailing blanks are removed from each line. Carriage control characters are translated to blank lines or ASCII form feed characters. If the file exists it will be overwritten.
Additional arguments may be specified after the file name:
sockdev
indicates the line printer is a socket device wherein the
filename is actually a socket specification instead of a
device filename. When used, there must only be one filename
specified in the form: port or host:port.
The device then accepts
remote connections on the given TCP/IP port,
and writes data to the socket instead of to a device
file. This allows automatic remote spooling of line printer
data. The sockdev option is mutually exclusive with all other
printer options (e.g. crlf, etc) and must be specified alone.
crlf
specifies, for ASCII files, that carriage return line feed
sequences are written at the end of each line.
If the crlf argument is not specified, then
line-feeds only are written at the end of each line.
noclear
specifies that the output file will not be cleared to zero
bytes when it is opened.
If the noclear argument is not specified, then
any previous content of the file is destroyed when the file
is opened for output.
fcbcheck
specifies that an attempt to skip to a FCB channel for which no line number has been set will cause the command to be rejected with a unit check. This is the default.
nofcbcheck
specifies that an attempt to skip to a FCB channel for which no
line number has been set will cause the next line of output to be
printed on the next line on the printer output. The opposite,
fcbcheck, is the default.
rawcc
specifies that printer output CCWs are not to be interpreted, but simply dumped in hex to the printer output file. This is useful for debugging. Default is to interpret printer CCWs normally.
fcb=argument
specifies an initial FCB image to use for this printer. The argument
may either consist of 12 numbers separated by commas (these are the
line numbers for channels 1 to 12), or of pairs of numbers in the
format nn:chan where chan is the
channel number and nn is the line number that the channel
corresponds to. Use 00 to leave the line number unset
for a channel (see fcbcheck above). The default is
fcb=1,7,13,19,25,31,37,43,63,49,55,61
which is equivalent to
fcb=1:1,7:2,13:3,19:4,25:5,31:6,37:7,43:8,63:9,49:10,55:11,61:12
browse | print
specifies whether the output should be optimized and cleaned up for
browsing, or optimized for printing. The default is browse.
index=nn
specifies 3211 indexing. Valid values are 0 to 31. The default is 0.
lpi=6|8
specifies vertical spacing of 6 or 8 lines per inch.
lpp=nn
specifies the number of lines per page. The default is 66.
If the filename begins with the vertical bar '|' pipe character, then it is removed and the remainder of the filename is interpreted as a command line (the name of a program or batch file followed by any necessary arguments) to which to "pipe" the printer output to. This is known as the "print-to-pipe" feature. All printer output is then sent to the piped program's stdin input, and all of the piped program's stdout and stderr output is piped back to Hercules for displaying on the hardware console.
If the "print-to-pipe" command line contains arguments, then quotes must be placed around the entire filename string including the vertical bar, for example:
000E 1403 "|/usr/bin/lpr -Phplj" crlf (for Unix)
000E 1403 "|c:\utils\pr -s -PLPT1:" crlf (for Windows)
The above example uses the pr program downloaded
from
http://www.atnetsend.net/computing/
If the "print-to-pipe" command line itself contains quotes, then the command line must be enclosed in apostrophes instead of quotes, for example:
000E 1403 '|"c:\Program Files\My Utils\pr" -s -PLPT1:' crlf
Tim Pinkawa has an example which shows how the print-to-pipe feature can be used to create output in PDF format: http://www.timpinkawa.net/hercules/prtspool.html
Five types of tape emulation are supported (see further below).
The only required parameter is the device filename. All other parameters are optional and must follow the filename. Use '*' (asterisk) for the filename to specify an empty (unmounted) tape drive. The specified file, if other than '*', must exist.
Additionally, if the file name starts with the '@' character (at sign), the file really describes a list of tape emulation files to be loaded in succession.
The syntax of each line is identical to the information that can be specified after the device type when the options are specified directly after the device type in the configuration file.
If the emulation file filename in the file list is the '*' (asterisk) character, then this specifies a set of options to be applied to all additional emulation files specified in the file list.
Parameters are appended in succession. In all cases, if the same parameter is specified more than once, the last instance takes precedence.
Therefore, it is possible to specify a set of parameters in the base configuration file, another set on a '*' line, and another set for each individual line. Parameters are then appended in that order: options specified on the base device statement itself first, followed by those options specified on the '*' statement, and finally those specified on each individual file list statement last. A SCSI tape device should not be given in a file list.
Refer to the distributed source-code's "README.TAPE" document for additional information regarding system and application programming for tape devices and instructions regarding use of the emulated ACF (Automatic Cartridge Feeder) and AUTOMOUNT features for virtual (non-SCSI) tape devices.
These are real tape drives attached to the host machine via a SCSI interface. Hercules emulation always makes the drive appear as a channel attached device such as 3420 or 3480, although the underlying physical drive may be any type of SCSI attached tape drive, including 4mm or 8mm DAT, DLT, SCSI attached 3480/3490 cartridge drives, and SCSI attached 3420 open reel tape drives.
Host-attached SCSI tapes are read and written using variable length EBCDIC blocks and filemarks exactly like a mainframe tape volume, and as a result can be freely used/exchanged on either (i.e. SCSI tapes created on a real mainframe can subsequently be read by Hercules just fine, and a SCSI tape created by Hercules can be subsequently read on a mainframe just fine, thus providing a convenient means of exchanging data between the two).
If you plan on using SCSI tapes with Hercules you might also be interested in the AUTO_SCSI_MOUNT configuration option.
The only required device statement parameter for SCSI attached tape
drives is the name of the device as it is known
by the host operating system,
usually "/dev/nst0" (for Linux or
Windows) or "\\.\Tape0" (for
Windows only), where '0' means tape drive number
0 (your first or only host-attached SCSI tape drive), '1' means your
second host-attached SCSI tape drive, etc.
Depending on what actual model of SCSI tape drive you actually have and how it behaves, you may need to specify one or more additional optional parameters for Hercules to provide proper emulation of the desired device type. For example: a Quantum 'DLT' (Digital Linear Tape) SCSI tape drive does not return nor use a block-id format compatible with 3480/3490 drives (it instead uses a full 32-bit block-id just like the model 3590 does). It also does not support the Erase Gap CCW at all.
Thus, in order to use, for example, a Quantum DLT drive with Hercules, you MUST specify some special additional options to prevent the Erase Gap command from being issued to the drive as well as to inform Hercules that the drive uses 32-bit block-ids.
Please note that the below options define how the actual SCSI hardware
device behaves, which is completely different from the way the emulated
device will appear to behave to your guest. That is to say, if you define
your tape drive to Hercules as a 3480 device, then Hercules will perform
3480 device type emulation such that the device appears to your guest o/s
as if it were a 3480 device. If the actual SCSI device behaves as
a 3590 device however (perhaps using/returning 32-bit block-ids instead
of the expected 22-bit format block-ids that 3480's use), then you MUST
specify the --blkid-32 special option on your Hercules device
statement so that Hercules's emulation logic can know that it needs to
translate 22-bit block-ids to 32-bit ones before sending them to the
actual SCSI hardware (and vice versa: to translate 32-bit block-ids from
the actual SCSI drive into 22-bit format block-ids that your guest expects
from a 3480 device).
As explained just above, certain model SCSI tape drives such as the Quantum DLT series may require special handling in order to provide the desired proper device type emulation. These special options are:
--no-erg
This option is intended to prevent issuance of the Erase Gap command to those SCSI tape drives which do not support it (such as the Quantum DLT series). It causes Hercules's device emulation logic to ignore any Erase Gap commands issued to the drive and to return immediate 'success' instead.
This option should only be used (specified) for drives such as the Quantum, which support switching from read mode to write mode in the middle of a data stream without the need of an intervening Erase Gap command. Specifying it for any other model SCSI drive may cause incorrect functioning as a result of the Erase Gap command not being issued to the actual SCSI hardware.
Check the manufacturer information for your particular model of SCSI-attached tape drive (and/or use Fish's "ftape" Windows utility) to determine whether or not this option is needed for your particular drive.
--blkid-32
This option indicates that your SCSI-attached tape drive only supports 32-bit block-ids (as used by 3590 drives) and not the 22-bit format used by 3480/3490 drives. You should only specify this option if you intend to define the drive as a model 3480 or 3490 device, and then only if your actual SCSI drive uses 32-bit block-ids of course. If you define your Hercules tape drive as a model 3590 device however, then this option is of course not needed since model 3590 drives are already presumed to use 32-bit block-ids.
Specifying this option on a 3480/3490 device statement will cause Hercules device emulation logic to automatically translate the actual SCSI tape drive's 32-bit block-id into 22-bit format before returning it back to the guest operating system (since that is the format it expects it to be in for a model 3480/3490 drive), and to translate the guest's 22-bit format block-id into 32-bit format before sending it to the actual SCSI hardware (since that is the format that the actual hardware requires it to be in).
--blkid-22
The complete opposite of the above --blkid-32 option.
These are read-only files which usually reside on CDROM.
OMA virtual tapes consist of one CDROM file corresponding
to each physical file of the emulated tape. An ASCII text
file called the tape descriptor file (TDF) specifies the
names of the files which make up the virtual tape.
The argument specifies the name of the tape descriptor
file (for example /cdrom/tapes/uaa196.tdf)
Each file on the virtual tape can be in one of three formats:
TEXT
TEXT files consist of variable length ASCII records delimited by carriage return line feed sequences at the end of each record. Each record is translated to EBCDIC and presented to the program as one physical tape block.
FIXED nnnnn
FIXED files consist of fixed length
EBCDIC blocks of the specified length
(nnnnn)
HEADERS
HEADERS files consist of variable length EBCDIC blocks. Each block is preceded by a 12-byte header.
If you have any IBM manuals in Bookmanager format on CDROM,
you can see some examples of TDF files in the
\TAPES directory on the CDROM.
These contain a complete tape in one file. AWSTAPE files
consist of variable length EBCDIC blocks. Each block is
preceded by a 6-byte header. Filemarks are represented by
a 6-byte header with no data. This is the same format as is
used by the P/390.
The argument specifies the location of the AWSTAPE file
(for example ickdsf.aws)
These contain a complete tape in one file. FakeTape files
consist of variable length EBCDIC blocks. Each block is
preceded by a 12-ASCII-hex-character header. Filemarks are represented by
a 12-character header with no data. The FakeTape format is
used by the Flex-ES system from Fundamental Software Inc (FSI).
The argument specifies the location of the FakeTape file
(for example ickdsf.fkt). Note: "FLEX-ES" and
"FakeTape" are trademarks of Fundamental Software, Inc.
These contain a complete tape in one file and have the same
structure as the AWSTAPE format with the added ability to have
compressed data.
The first argument specifies the location of the HET file. The
filename must end with ".het" to be recognized by Hercules as an
HET file.
(for example 023178.het)
Additional arguments that allow you to control various HET settings are:
AWSTAPE
The AWSTAPE argument causes HET files to
be written in AWSTAPE format. This basically, disables
the additional features provided by the HET format.
COMPRESS=n
IDRC=n
COMPRESS and IDRC control
whether compression should be used when writing to HET
files. The value n can be 1
to turn on compression (the default) or 0 to turn
it off. IDRC is currently a synonym for
COMPRESS, but may be used in the future to
control other emulated tape drive features.
METHOD=n
The METHOD option allows you to specify
which compression method to use. You may specify
1 for ZLIB compression or 2
for BZIP2 compression. The default is 1.
LEVEL=n
The LEVEL option controls the level of
compression. It ranges from 1 for fastest
compression to 9 for best compression.
The default is 4.
CHUNKSIZE=nnnnn
The CHUNKSIZE option allows you to create
HET files that contain different chunk sizes. The AWSTAPE
(and therefore the HET) format allows each tape block to be
logically broken up into smaller chunks. For instance, if
your S/3x0 application creates tapes with a block size of
27998, those blocks would be broken down into
nnnnn sized chunks.
The range is from 4096
to 65535, the latter being the default.
Decreasing the value from its default may reduce compression
performance.
For compatability with AWSTAPE files created by the P/390,
specify AWSTAPE with CHUNKSIZE=4096.
The following parameters apply to AWS, HET and FakeTape emulation files:
MAXSIZE=n | MAXSIZEK=n
| MAXSIZEM=n
Specifies the maximum size (in bytes, Kilobytes or Megabytes) that the emulated file is allowed to grow to. Specifying zero for this parameter means "unlimited" (i.e. there is no limit).
EOTMARGIN=n
Specifies the number of bytes remaining before reaching maxsize at which point the tape device will signal the presence of the "End of Tape" marker (reflector), thus allowing the program to switch to the next tape.
READONLY=n
Specifies whether the tape is mounted read-only (without a write ring or with the cartridge protect switch set to "write protect"). A parameter of 1 means read-only; a parameter of 0 means read-write. If READONLY=1, RO or NORING is not specified, the default is READONLY=0. Note that READONLY=0 does not override the host system file permission settings for the underlying AWS or HET file. If the AWS or HET file is marked read-only, the tape will be mounted read-only despite specification of READONLY=0.
RO
NORING
Specifies that the tape is mounted read-only (without a write ring or with the cartridge protect switch set to "write protect"). RO and NORING are equivalent to READONLY=1.
RW
RING
Specifies that the tape should be mounted read-write, if possible. RW and RING are equivalent to READONLY=0. This is the default if RO, NORING or READONLY=1 is not specified. Note that RW and RING do not override the host system file permission settings for the underlying AWS or HET file. If the AWS or HET file is marked read-only, the tape will be mounted read-only despite specification of RW or RING.
DEONIRQ=n
Specifies whether a device end is presented if intervention is required during tape motion. A parameter of 1 selects this option; a parameter of 0 turns it off.
NOAUTOMOUNT
Indicates support for guest-initiated automatic tape volume mounting is to always be disabled for this tape device.
Automatic guest tape-mount support is automatically globally
enabled for all virtual (non-SCSI) tape devices by default
whenever an allowable automount directory is defined via the
AUTOMOUNT configuration file statement
or the automount panel command.
The NOAUTOMOUNT option allows you to specifically
disable such support for a given device.
The automount feature enables software running in guest operating
systems to automatically mount, unmount and/or query for themselves
the host "virtual tape volume" filename mounted on a tape drive,
via the use of special CCW opcodes (0x4B Set Diagnose and 0xE4
Sense Id) without any intervention on the part of the Hercules
operator. An example of such a program for DOS/VSE called
TMOUNT is provided in the util
subdirectory of the distributed source code.
This is a sticky option. When specified, automount support for
the device remains disabled until the option is specifically
removed via a devinit command without the option
specified. This means if NOAUTOMOUNT is enabled
for a device while global automount functionality is currently
disabled (because no AUTOMOUNT statement
was specified at Hercules startup), then automount functionality
remains disabled for the device even should global automount
functionality be later manually enabled via an
automount panel command.
When the 0x4B Set Diagnose CCW is used to auto-mount a virtual tape volume onto a given tape drive, an absolute (fully-qualified) pathname should normally always be specified, but need not be if a path relative to the currently defined "default allowable" automount directory is used instead.
The default allowable automount directory is always the first "allowable" directory that was defined, or else the current directory if no allowable directories were specifically defined. (There is always a default allowable directory whenever any allowable or unallowable automount directories are defined.)
Fully-resolved, absolute-full-path filenames are defined as being those which, for Windows, have a ':' (colon) in the second position or, for other host operating systems (e.g. Linux), have a '/' (slash) in the first position. Paths which start with a '.' (period) are considered relative paths and will always be appended to the currently defined default allowable automount directory, before being resolved into fully-qualified paths by the host system. (I.e. only fully-resolved absolute pathnames are used in the performance of the actual automatic tape volume mount.)
For example, if more than one allowable automount directory is defined and the volume wishing to be mounted happens to reside in the second one, then a fully-qualified absolute pathname should of course be specified (or else one that is relative to the default directory which happens to resolve to the desired file).
All attempts to automount host files in a "disallowed" directory or any of its subdirectories will be rejected. Similarly any attempt to automount a file which is not within any "allowable" directory or subdirectory will be rejected. An error message is always issued in such cases. A message is also issued whenever a successful mount or unmount is performed.
A sample guest automount program called TMOUNT for
the DOS/VSE operating system is provided in the
util subdirectory of the distributed source code.
The first argument defines the emulation type, and the remaining arguments depend on the chosen emulation type. If the first argument is not a recognized emulation type, then the driver will operate as in Hercules Version 1, using Willem Konynenberg's vmnet package, as described in Axel Schwarzer's CTCA 3088 document.
The following are the emulation types currently supported:
A point-to-point IP connection with the TCP/IP stack of the driving system on which Hercules is running. See the Hercules TCP/IP page for details.
(Note: The CTCI protocol is only for the Linux version of Hercules. For Windows, use the below CTCI protocol instead).
A modified Win32 version of the CTCI protocol for the Windows crowd. Note that the protocol name (CTCI) is the same, even though the actual implementation is very different. See Fish's CTCI-W32 page for further details and information.
guestip
specifies the IP address of the guest operating system running under Hercules.
hostip
specifies the IP address of the host (Linux or Windows) side of the point-to-point link. This may or may not be the same as your system's regular IP address. For Windows, if the host system is configured with DHCP, this should instead be the MAC address of the Ethernet adapter you wish to use to have Hercules communicate with the outside world.
If these arguments are specified, they must precede the required arguments.
-k kernel-capture-buffer-size
-i tuntap32-i/o-buffer-size
See Fish's CTCI-W32 page for further details and information.
If these arguments are specified, they must precede the required arguments:
-n name or --dev name
specifies the name of the tunnel device to use.
The default for Linux is /dev/net/tun (which is correct for version 2.4 and above of the Linux kernel).
For Windows, specify the IP address or MAC address of the real Windows adapter to emulate the virtual guest's adapter on. The default is the first adapter found according to Windows' adapter binding order, which may not be the one you want if you have multiple adapters.
See question #22 of the CTCI-W32 F.A.Q. (Frequently Asked Questions) document for more information about adapter binding order, and the CTCI-W32 Configuration web page for general information regarding CTCI-W32 configuration.
-m MAC address or --mac MAC address
where 'MAC address' is the optional hardware address for
the virtual interface in the format: hh:hh:hh:hh:hh:hh. The default value
is '00:00:5E:nn:nn:nn' where the :nn:nn:nn portion is constructed
from the last 3 octets of the specified guestip.
-s netmask
where netmask is the netmask to use for the automatically added point-to-point route in standard dotted internet noitation (e.g. 255.255.255.0)
-d or --debug
specifies that debugging output is to be produced on the Hercules control panel. This should normally be left unspecified.
An emulated CTCA to another Hercules system. This emulation mode appears to the operating system running in the Hercules machine as an IBM 3088 Channel to Channel Adapter. It provides communication via a TCP connection with another instance of the CTCT driver, and is designed to carry TCP/IP communications between two guest TCP/IP stacks. CTCT may also be used for communication between the client and server components of the MVS Dynamic Debug Tool.
Four arguments are required:
lport
specifies the local TCP port. This is the TCP port that Hercules will listen on for this CTCA.
rhost
specifies the remote host. This is the name or IP address of the remote system that Hercules is running on, not the name or IP address of the OS running on that copy of Hercules.
rport
specifies the remote TCP port. The rport parameter on this system must match the lport parameter on the remote system, and vice versa.
bufsize
specifies the buffer size for the link. If this link is used for IP traffic, this parameter should be more than the MTU of the interface definition in the OS.
Note: CTCT only supports IP traffic. Use CTCE to transport general purpose payloads such as NJE, SNA, PVM, etc.
The CTCE device type will emulate a real 3088 Channel to Channnel Adapter also for non-IP traffic, enhancing the CTCT capabilities. CTCE connections are also based on TCP/IP between two (or more) Hercules instances, and requires an even-odd pair of port numbers per device side. Only the even port numbers are to be configured; the odd numbers are just derived by adding 1 to the (configured) even port numbers. The socket connection pairs cross-connect, the arrows showing the send->receive direction :
x-lport-even -> y-rport-odd
x-lport-odd <- y-rport-even
Three arguments are required:
lport
raddress
rport
The remaining arguments are optional:
mtu
sml
A sample CTCE device configuration is shown below:
Hercules PC Host A with IP address 192.168.1.100 :
0E40 CTCE 30880 192.168.1.200 30880
0E41 CTCE 30882 192.168.1.200 30882
Hercules PC Host B with IP address 192.168.1.200 :
0E40 CTCE 30880 192.168.1.100 30880
0E41 CTCE 30882 192.168.1.100 30882
CTCE connected Hercules instances can be hosted on either Unix or Windows platforms, both sides do not need to be the same.
An emulated Lan Channel Station Adapter. This emulation mode appears to the operating system running in the Hercules machine as an IBM 8232 LCS device, an IBM 2216 router, a 3172 running ICP (Interconnect Communications Program), the LCS3172 driver of a P/390, or an IBM Open Systems Adapter.
Rather than a point-to-point link, this emulation creates a virtual ethernet adapter through which the guest operating system running in the Hercules machine can communicate. As such, this mode is not limited to TCP/IP traffic, but in fact will handle any ethernet frame.
The configuration statement for LCS is as follows:
NOTE: There are no required parameters for the LCS emulation, however there are several options that can be specified on the config statement:
NOTE: On the MAC OS X Platform, the long option format (--xxx) is not supported. Only the short option format (-x : one dash, one letter) should be used.
-n devname or --dev devname
where devname is:
the name of the TUN/TAP special character device, normally /dev/net/tun.
is either the IP or MAC address of the driving systems network card. TunTap32 will automatically select the first network card it finds if this option is omitted, this may not be desirable for some users.
-o filename or --oat filename
where filename specifies the filename of the Address
Translation file. If this option is specified, the optional
--mac and guestip entries are ignored in preference to
statements in the OAT. (See further below for the syntax
of the OAT file)
-m MAC Address or --mac MAC address
where MAC Address is the optional hardware address of
the interface in the format: hh:hh:hh:hh:hh:hh. If you use the
--oat option, do not specify an address here.
guestip
is an optional IP address of the Hercules
(guest OS) side. Note: This is only used to
establish a point-to-point routing table entry
on driving system. If you use the --oat option,
do not specify an address here.
The syntax for the Address Translation file is as follows:
|
Dev
Mode
Note: the SNA operation mode is NOT currently implemented.
Port
For IP modes, the entry specific information is as follows:
PRI | SEC | NO
specifies where a packet with an unknown IP address is forwarded to. PRI is the primary default entry, SEC specifies the entry to use when the primary is not available, and NO specifies that this is not a default entry.
nnn.nnn.nnn.nnn
specifies the home IP address
When the operation mode is IP, specify only the even (read) device number dev. The odd (write) address will be create automatically.
Note: the SNA operation mode is NOT currently implemented.
Additionally, two other statements may be included in the address translation file. The HWADD and ROUTE statements:
HWADD pp hh:hh:hh:hh:hh:hh
Use the HWADD to specify a hardware (MAC) address for a virtual adapter. The first parameter after HWADD specifies with relative adapter for which the address is applied.
ROUTE pp nnn.nnn.nnn.nnn ...
The ROUTE statement is included for convenience. This allows the hercifc program to create a network route for this specified virtual adapter. Please note that it is not necessary to include point-to-point routes for each IP address in the table. This is done automatically by the emulation module.
Up to 4 virtual (relative) adapters 00-03 are currently supported.
If no Address Translation file is specified, the emulation module will create the following:
The argument specifies the name of a file containing the disk CKD DASD image or the INET address of a Hercules shared device server.
The file consists of a 512-byte device header record followed by fixed length track images. The length of each track image depends on the emulated device type, and is always rounded up to the next multiple of 512 bytes.
Volumes larger than 2GB (for example, the 3390 model 3) can be supported by spreading the data across more than one file. Each file contains a whole number of cylinders. The first file (which contains cylinders 0-2518 in the case of a 3390) usually has _1 as the last two characters of its name. The ckddasd driver allocates the remaining files by replacing the last character of the file name by the characters 2, 3, etc.
Note: When CKD DASD images are spread across multiple files, you must specify only the first file name (the file with suffix _1) in the configuration statement.
If your operating system supports large file sizes (or 64-bit offsets) then volumes larger than 2G can be kept in a single file.
Alternatively, the argument may specify the name of a file containing a compressed CKD DASD image. The CKD driver will automatically detect whether the file contains a regular CKD image or a compressed CKD image.
Refer to "Creating an empty DASD volume" in the "Creating, formatting, and loading DASD volumes" section of the Creating DASD web page for information on using the 'dasdinit' command/utility to create compressed dasd files. Refer to the Compressed Dasd Emulation page for details on the actual CCKD emulation itself and additional information on the CCKD initialization/tuning control file statement.
If you specify an INET address, the format is:
ip-name-or-addr:port:devnum
ip-name-or-addr specifies the internet name or address where the Hercules shared device server is running.
port specifies the port number the shared device server is listening on. If omitted, the default is 3990.
devnum specifies the device number on the shared device server. If omitted, the default is the current device number.
In addition to the above, some additional optional arguments are also supported.
sf=shadow-file-name
A shadow file contains all the changes made to the emulated dasd since it was created, until the next shadow file is created. The moment of the shadow file's creation can be thought of as a snapshot of the current emulated dasd at that time, because if the shadow file is later removed, then the emulated dasd reverts back to the state it was at when the snapshot was taken.
Using shadow files, you can keep the base file on a read-only device such as cdrom, or change the base file attributes to read-only, ensuring that this file can never be corrupted.
Hercules console commands are provided to add a new shadow file, remove the current shadow file (with or without backward merge), compress the current shadow file, and display the shadow file status and statistics
For detailed information regarding shadow files and their use, please see the "Shadow Files" section of the Compressed Dasd Emulation web page.
[no]syncio
syncio enables possible 'synchronous' i/o. This is a dasd i/o feature wherein guest i/o requests are completed "synchronously" during the actual emulated execution of the SIO/SSCH (start-i/o / start subchannel) instruction rather than being deferred and executed asynchronously in a separate device i/o thread.
Only i/o which are known to be able to be completed without actually needing to perform any actual host i/o are completed synchronously (e.g. whenever the data being requested is found to already be in cache). If the requested data is not found in the cache, then an actual host i/o will need to be done and the request is passed to a device i/o thread to be completed asyncronously instead.
syncio is the default for ckd. syncio statistics may be displayed via
the Hercules syncio panel command.
syncio may be abbreviated as
syio
readonly
readonly returns "write inhibited" sense when a write is attempted. Note that not all of the sense bits may be getting set absolutely correctly however. (Some people have reported getting different error messages under Hercules than a real machine, but it really hasn't been an issue for a while now.)
readonly may be abbreviated as
rdonly or ro
fakewrite
fakewrite is a kludge for the readonly sense problem mentioned above. Here the disk is not intended to be updated (MVS updates the DSCB last referenced field for a readonly file) and any writes appear to be successful even though nothing actually gets written.
fakewrite may be abbreviated as
fakewrt or fw
[no]lazywrite
[no]fulltrackio
These options have been deprecated. They are still accepted, but they do absolutely nothing.
fulltrackio may be abbreviated as
fulltrkio or ftio
cu=type
Specifies the type of control unit to which this device is attached. The use of this parameter does not necessarily imply that all functions of the specified control unit are emulated, its only purpose is to force a particular control unit type to be indicated in the data returned by SENSE ID and similar CCW's.
The default value depends on the device type:
| Device type | Default CU type |
|---|---|
| 2311 | 2841 |
| 2314 | 2314 |
| 3330 3340 3350 3375 3380 | 3880 |
| 3390 | 3990 |
| 9345 | 9343 |
Other values which may be specified are: 3990-3 and 3990-6.
Normally the default value is appropriate and this parameter need not be specified.
The argument specifies the name of a file which contains the FBA DASD image or the INET address of a Hercules shared device server.
The file consists of fixed length 512-byte records, each of which represents one physical block of the emulated disk.
To allow access to a minidisk within a full-pack FBA DASD image file, two additional arguments may be specified after the file name:
origin
specifies the relative block number within the DASD image file at which the minidisk begins. The number must be less than the number of blocks in the file. The default origin is zero.
numblks
specifies the number of 512-byte blocks in the minidisk. This number must not exceed the number of blocks in the file minus the origin. If omitted, or if specified as an asterisk, then the minidisk continues to the end of the DASD image file.
If you specify an INET address the format is:
ip-name-or-addr:port:devnum
ip-name-or-addr specifies the internet name or address where the Hercules shared device server is running.
port specifies the port number the shared device server is listening on. If omitted, the default is 3990.
devnum specifies the device number on the shared device server. If omitted, the default is the current device number.
In addition to the above, some additional optional arguments are also supported.
sf=shadow-file-name
The handling of shadow files for FBA devices is identical as that for
CKD devices. Please refer to the preceding CKD section for information
regarding use of the sf= shadow file option.
[no]syncio
syncio enables possible 'synchronous' i/o and is explained in detail in the preceding CKD dasd section. Note however that syncio is currently disabled by default for FBA dasd due to an as yet unresolved problem and must therefore be specifically enabled if you wish to use it for FBA dasd.
syncio may be abbreviated as
syio
( Preliminary 2703 BSC Support )
Describes a BSC emulation line entry to either link 2 Hercules engines or a custom made program emulating a 2780, 3780 or 3x74, or a custom made program interfacing to a real BSC line.
The communication is emulated over a TCP connection. All bytes are transfered as-is (except for doubling DLE in transparent mode) just like it would over a real BSC link. Emulated EIA (DCD, DTR, CTS, etc..) or X.21/V.11 leads (C, T, etc..) are treated differently depending on the DIAL option selected.
The line emulates a point-to-point BSC link. There is no point-to-multipoint handling.
The following options define the line emulation behaviour:
DIAL=IN | OUT | INOUT | NO
Specifies call direction (if any). If DIAL=NO is specified, the
TCP outgoing connection is attempted as soon as an 'ENABLE' CCW is executed.
Also, in this mode, an incoming connection will always be accepted. If DIAL=IN|INOUT
is specified, a TCP incoming call is accepted ONLY if an 'ENABLE' CCW is currently
executing on the device. If DIAL=OUT, the 'ENABLE' CCW is rejected.
When DIAL=IN|INOUT is specified, a DIAL CCW allows the application
to establish a TCP connection to a specific host. For other DIAL values,
the DIAL CCW is rejected.
lhost=hostname | ip address | *
Specifies which IP address to listen on. This also conditions the network
interface from which incoming calls will be accepted. Specifying '*' means
all incoming TCP calls are accepted, regardless of the destination IP
address or call origin. This is the default value. Specifying a specific
IP address when DIAL=OUT is specified has no effect.
lport=service name | port number
Specifies the TCP port for which to listen to incoming TCP calls. This
value is mandatory for DIAL=IN|INOUT|NO. It is ignored for DIAL=OUT.
rhost=hostname | ip address
rport=service name | port number
Specifies the remote host and port to which to direct a TCP connection on a
DIAL=NO line when an 'ENABLE' CCW is executed. This option is mandatory when DIAL=NO
is specified. It is ignored for other DIAL values.
rto=0 | -1 | nnn | 3000
Specifies the number of milliseconds before terminating a read on a timeout, when no read termination control character is received. Specifying 0 means the READ ends immediately. -1 specifies there is no timeout.
pto=0 | -1 | nnn | 3000
Specifies the number of milliseconds before terminating a POLL on a timeout, when no ACK or NACK sequence is received. Specifying 0 means the POLL ends immediately. -1 specifies there is no timeout.
eto=0 | -1 | nnn | 10000
Specifies the number of milliseconds before terminating an ENABLE operation on a timeout.
the timeout applies when DIAL=NO|IN|INOUT is specified, the outgoing TCP call
fails (DIAL=NO) and there is no previously or currently established TCP connection
for this line. When DIAL=NO is specified, the timeout defaults to 10 seconds.
For DIAL=IN|INOUT, the timeout defaults to -1.
( Preliminary 2703 TELE2 TTY Support )
Describes a 2703 Telegraph Terminal Control Type II (TTY 33/35) stop/start line, providing access to the Host OS via a standard TELNET client.
To the host OS the line emulates an asynchronous TELE2 connection. The communication is emulated over a TELNET connection.
The following options define the line emulation behaviour:
lport=port number
Specifies the TCPIP port to listen on for incoming TCP calls.
dial=IN
Specifies that this line is for in-bound calls. Required.
tty=1
Specifies that this definition is for a TTY port. Required
If you have a question about Hercules, see the Hercules Frequently-Asked Questions page.
Last updated $Date$ $Revision$