HP-UX Reference (11i v2 07/12) - 1M System Administration Commands N-Z (vol 4)
s
swinstall(1M) swinstall(1M)
If set to false, SD does not compute checksums and compares files only by size and times-
tamp.
remove_obsolete_filesets=false
(Applies only to swcopy.) Controls whether
swcopy automatically removes obsolete
filesets from target products in the target depot. If set to
true, swcopy removes obsolete
filesets from the target products that were written to during the copy process. Removal
occurs after the copy is complete. Filesets are defined as obsolete if they were not part of
the most recent packaging of the product residing on the source depot.
retry_rpc=1
Defines the number of times a lost source connection is retried during file transfers in
swinstall or swcopy. A lost connection is one that has timed out. When used in con-
junction with the rpc_timeout option, the success of installing over slow or busy net-
works can be increased. If set to zero, any
rpc_timeout to the source causes the task
to abort. If set from 1 to 9, the install of each fileset is attempted that number of times.
The
reinstall_files
option should also be set to false to avoid installing files within
the fileset that were successfully installed.
This option also applies to the controller contacting the agent. If the agent session fails to
start for any reason, the controller tries to recontact that agent for the number of times
specified in
retry_rpc, using the values from the retry_rpc_interval
option to
determine how long to wait between each attempt to recontact the agent.
retry_rpc_interval={0}
Specifies in minutes the length of the interval for repeated attempts to make a connection
to a target after an initial failure. Used in conjunction with the
retry_rpc option. If
the number of values in this option equals the value of retry_rpc, SD tries reestablish-
ing a source connection for the number of times specified in retry_rpc. If the number of
values in retry_rpc_interval
is less than the value in retry_rpc, SD repeats the
final interval value until the number of retries matches
retry_rpc.
For example, if a session failed to start and retry_rpc was set to 9 and
retry_rpc_interval
was set to {1 2 4 8 15} to allow long waits to handle transient
network failures, the SD controller would attempt to recontact the agent after 1 minute for
the first retry, then 2 minutes for the second retry, 4 for the third, then 8, then 15 for all
additional retries until nine retries were attempted. With these values, a file load failure
could cause the operation to pause for 90 minutes (1+2+4+8+15+15+15+15+15). If
retry_rpc was set to 5 and retry_rpc_interval
was set to {1 2 4 8 15}, the con-
troller would try to contact the target five times over a 30-minute period.
rpc_binding_info=ncacn_ip_tcp:[2121] ncadg_ip_udp:[2121]
Defines the protocol sequence(s) and endpoint(s) on which the daemon listens and the other
commands contact the daemon. If the connection fails for one protocol sequence, the next
is attempted. SD supports both the tcp (
ncacn_ip_tcp:[2121]
) and udp
(
ncadg_ip_udp:[2121]
) protocol sequence on most platforms. See the sd(5) man
page by typing
man 5 sd for more information.
rpc_binding_info_source=
Defines the protocol sequence(s) and endpoint(s) on which commands contact the daemon
for source access only. If the connection fails for one protocol sequence, the next is
attempted. If this is set to no value (default) the values from
rpc_binding_info are
used to contact the daemon for source access. See
rpc_binding_info (above) for more
information.
rpc_binding_info_target=
Defines the protocol sequence(s) and endpoint(s) on which commands contact the daemon
for target access only. If the connection fails for one protocol sequence, the next is
attempted. If this is set to no value (default) the values from rpc_binding_info are
used to contact the daemon for target access. See
rpc_binding_info (above) for more
information.
rpc_timeout=5.
Relative length of the communications timeout. This is a value in the range from 0 to 9 and
is interpreted by the DCE RPC. Higher values mean longer times; you may need a higher
value for a slow or busy network. Lower values give faster recognition on attempts to con-
tact hosts that are not up or not running swagentd. Each value is approximately twice
HP-UX 11i Version 2: December 2007 Update − 14 − Hewlett-Packard Company 465