Compaq ACMS for OpenVMS
Remote Systems Management
Guide
8.6 Interfaces Table
The Interfaces table is populated from the configuration file by the
ACMS Remote Manager process during process startup. This table
specifies which interfaces are active and contains parameters
associated with each interface. By default, the RPC interface is
started; SNMP is not started .
Table 8-5 Interfaces Table
Field |
Data Type |
SNMP Access |
RPC Access |
Configuration Access |
Dynamic |
interface
|
string
|
R
|
R
|
R
|
|
state
|
state1
|
RW
|
RW
|
RW
|
D
|
running_state
|
state2
|
R
|
R
|
R
|
|
get_request_count
|
integer
|
R
|
R
|
|
|
set_request_count
|
integer
|
R
|
R
|
|
|
alarms_sent
|
integer
|
R
|
R
|
|
|
time_alarm_last_sent
|
integer
|
R
|
R
|
|
|
Key to Access Modes
R---Read Access
RW---Read/Write Access
Blank---Not available to the interface
D---Field is dynamic.
8.6.1 Field Descriptions
Following are descriptions of the fields in Table 8-5.
- interface
Name or type of the interface. Valid values are RPC
or SNMP.
- state
Current state as configured, either from the
configuration file or by a user at run time. Valid values are ENABLED
or DISABLED. Note that state and acms_state are not always the same
because of potential run-time failures in a thread. For instance, if a
thread fails to start, state may be ENABLED, but acms_state may be
STOPPED.
A thread can be enabled only if the acms_state value is
STOPPED. A thread can be disabled only if the acms_state value is not
STOPPED.
- acms_state
Actual execution state. Interfaces go through the
following states:
- INITING
The Remote Manager is in the process of creating the
interface thread.
- STARTING
The interface thread has been created and is
initializing.
- STARTED
The interface thread has completed initializing and is
now running.
- STOPPING
The thread is starting shutdown, as the result of
either a stop request or a fatal error.
- STOPPED
The thread is no longer executing.
- get_request_count
The number of read requests submitted to the
interface. This includes requests that are rejected because of
authorization failures.
- set_request_count
The number of write requests submitted to
the interface. This includes requests that are rejected because of
authorization failures.
- alarms_sent
The number of alarms that have been sent by this
interface. For SNMP, these are SNMP traps. For RPC, this field is
undefined.
- time_alarm_last_sent
The time the most recent alarm was sent by
this interface. For SNMP, this is the time the last SNMP trap was sent.
For RPC, this field is undefined.
8.7 Manager Status Table
The Manager Status table contains run-time values that reflect Remote
Manager activity. This table is maintained internally by the Remote
Manager and is read only to all external entities. Values in the table
can be accessed through one of the supported interfaces. No changes can
be made to the table by external users.
In general, the values in this table are informational only.
Table 8-6 Manager Status Table
Field Name |
Data Type |
SNMP Access |
RPC Access |
Dynamic |
collection_count
|
integer
|
R
|
R
|
|
interfaces_count
|
integer
|
R
|
R
|
|
timer_count
|
integer
|
R
|
R
|
|
trap_count
|
integer
|
R
|
R
|
|
rpc_udp_state
|
state1
|
R
|
R
|
|
rpc_tcp_state
|
state1
|
R
|
R
|
|
Key to Access Modes
R---Read Access
RW---Read/Write Access
Blank---Not available to the interface
D---Field is dynamic
8.7.1 Field Descriptions
Following are descriptions of the fields in Table 8-6.
- collection_count
Current number of Collection table entries.
- interfaces_count
Current number of entries in the Interfaces
table.
- timer_count
Current number of entries in the Timer table.
- trap_count
Current number of entries in the Trap table.
- rpc_udp_state
Current state of the RPC interface using the UDP
protocol. A value of 1 means that the UDP protocol is active. A value
of 0 means that the UDP protocol is inactive.
- rpc_tcp_state
Current state of the RPC interface using the TCP
protocol. A value of 1 means that the UDP protocol is active. A value
of 0 means that the UDP protocol is inactive.
8.8 Parameter Table
The Parameter table contains values that control the operation of the
ACMS Remote Manager and that are not directly related to any ACMS
entity. This table is populated initially from the ACMSCFG file. The
Remote Manager maintains the table internally at run time; users can
access data in the table only through one of the supported interfaces.
Changes made to the table at run time are lost when the Remote Manager
is stopped.
In general, the values in this table should be modified for fine tuning
only, and only if a demonstrated need exists.
Note
All the fields in Table 8-7 are of type integer, and all fields
have read and write access.
|
Table 8-7 Parameter Table
Field |
Default Value |
Minimum Value |
Maximum Value |
Dynamic |
Interface |
dcl_audit_level
|
E
|
0
|
F
|
D
|
S,R,F
|
dcl_mgr_priority
|
5
|
1
|
10
|
|
S,R,F
|
dcl_stacksize
|
300
|
1
|
2147483647
|
|
S,R,F
|
event_log_priority
|
5
|
1
|
10
|
|
S,R,F
|
log_stacksize
|
300
|
1
|
2147483647
|
|
S,R,F
|
login_creds_lifetime
|
60
|
1
|
14399999
|
D
|
S,R,F
|
max_logins
|
20
|
1
|
2147483647
|
D
|
S,R,F
|
max_rpc_return_recs
|
20
|
1
|
2147483647
|
|
S,R,F
|
mgr_audit_level
|
E
|
0
|
F
|
D
|
S,R,F
|
msg_proc_audit_level
|
E
|
0
|
F
|
D
|
S,R,F
|
msg_proc_priority
|
5
|
1
|
10
|
|
S,R,F
|
msg_proc_stacksize
|
300
|
1
|
2147483647
|
|
S,R,F
|
mss_coll_interval
|
10
|
1
|
863999999
|
D
|
S,R,F
|
proc_mon_audit_level
|
E
|
0
|
F
|
|
S,R,F
|
proc_mon_interval
|
30
|
1
|
14399999
|
D
|
S,R,F
|
proc_mon_priority
|
5
|
1
|
10
|
|
S,R,F
|
proc_mon_stacksize
|
300
|
1
|
2147483647
|
|
S,R,F
|
proxy_creds_lifetime
|
60
|
1
|
14399999
|
D
|
S,R,F
|
rpc_audit_level
|
E
|
0
|
F
|
D
|
S,R,F
|
rpc_priority
|
5
|
1
|
10
|
|
S,R,F
|
rpc_stacksize
|
300
|
1
|
2147483647
|
|
S,R,F
|
security_audit_level
|
E
|
0
|
F
|
D
|
S,R,F
|
snmp_agent_time_out
|
10
|
1
|
863999999
|
D
|
S,R,F
|
snmp_audit_level
|
E
|
0
|
F
|
D
|
S,R,F
|
snmp_are_you_there
|
300
|
1
|
863999999
|
|
S,R,F
|
snmp_priority
|
5
|
1
|
10
|
|
S,R,F
|
snmp_sel_time_out
|
5
|
0
|
863999999
|
|
S,R,F
|
snmp_stacksize
|
300
|
1
|
2147483647
|
|
S,R,F
|
timer_audit_level
|
E
|
0
|
F
|
D
|
S,R,F
|
timer_interval
|
30
|
1
|
863999999
|
D
|
S,R,F
|
timer_priority
|
5
|
1
|
10
|
|
S,R,F
|
timer_stacksize
|
300
|
1
|
2147483647
|
|
S,R,F
|
total_entity_slots
|
20
|
1
|
2147483647
|
|
S,R,F
|
trace_msg_wait_time
|
5
|
1
|
14399999
|
D
|
S,R,F
|
trace_start_wait_time
|
5
|
1
|
14399999
|
D
|
S,R,F
|
trap_audit_level
|
E
|
0
|
F
|
D
|
S,R,F
|
trap_priority
|
5
|
1
|
10
|
|
S,R,F
|
trap_stacksize
|
300
|
1
|
2147483647
|
|
S,R,F
|
wksp_coll_interval
|
10
|
1
|
863999999
|
D
|
S,R,F
|
Key to Interface
S---SNMP
R---RPC (API and ACMSMGR utility)
F---File (configuration file)
D---Field is dynamic.
8.8.1 Field Descriptions
Following are descriptions of the fields in Table 8-7.
- dcl_audit_level, mgr_audit_level,
msg_audit_level,proc_mon_audit_level, rpc_audit_level,
security_audit_level, snmp_audit_level, timer_audit_level,
trap_audit_level
Audit levels determine the amount of auditing
information written for a given facility. Audit levels are specified
using a hexidecimal value from 0 (none) to F (all). The integer values
are a logical ORing of the following:
INFO
|
1
|
WARN
|
2
|
ERROR
|
4
|
FATAL
|
8
|
For example, to specify auditing of both error and fatal
information, specify a value of C. For more information about auditing
and audit levels see Section 3.7.
- dcl_mgr_priority
Relative priority of the DCL manager thread.
The DCL manager is used to send ACMS run-time changes to the ACMS
system. Priority is specified as a whole number between 1 and 10, where
1 is the lowest priority and 10 is the highest. This value should be
left at the default.
- dcl_stacksize, log_stacksize, msg_proc_stacksize,
proc_mon_stacksize, rpc_stacksize, snmp_stacksize, timer_stacksize,
trap_stacksize,
These values determine the internal stack sizes for
each thread. Stack sizes are set during thread creation and are not
adjusted after the thread has been started. Restartable threads, such
as RPC and SNMP, can be adjusted while the Remote Manager is running by
disabling the interface, modifying the parameter, and then reenabling
the interface.
- event_log_priority
Relative priority of the event log thread.
The event log thread writes audit messages to the audit log. Priority
is specified as a whole number between 1 and 10, where 1 is the lowest
priority and 10 is the highest. This value should be left at the
default.
- login_creds_lifetime
The amount of time (in minutes) that
explict logins are valid. When a user logs in to a Remote Manager
process using a valid OpenVMS account and password, a login is created
for the user, and the expiration of that login is calculated and stored
based on this parameter. When the current time is greater than the
expiration time, the user is logged out and must log in again using the
ACMSMGR LOGIN command. A change to this parameter takes effect for any
login that takes place after the change is made. A change to this
parameter does not take effect for any login that took place before the
change was made.
- max_logins
Maximum number of external processes allowed to
concurrently connect to the Remote Manager. Starting the SNMP interface
counts as one login. Each RPC client counts as one login. RPCs are
serviced serially.
- max_rpc_return_recs
The maximum number of records to be
returned to any given request for data. This parameter allows network
bandwidth to be conserved by sending data in user-managed chunks.
- msg_proc_priority
Relative priority of the message processor
thread. The message processor is responsible for removing messages sent
by ACMS processes to the Remote Manager from the error input queue and
for processing messages according to configuration values specified in
the Collection and Trap tables. This value should be left at the
default. Priority is specified as a whole number between 1 and 10,
where 1 is the lowest priority and 10 is the highest.
- mss_coll_interval
Controls the frequency (in seconds) at which
MSS values are collected. A lower value causes MSS values to be
collected more often; a higher value causes MSS values to be collected
less often. MSS values are collected by all ACMS run-time processes
except SWL, ATR, and procedure servers.
- proc_mon_interval
The frequency (in seconds) at which the
process monitor thread should run. The process monitor thread checks
for the existence of the ACC and other ACMS run-time processes in order
to map the MGMT global section and to send alarms.
- proc_mon_priority
Relative priority of the process monitor
thread. The process monitor thread periodically checks for the
existence of the ACC process in order to map the MGMT global section
and to send alarms. Priority is specified as a whole number between 1
and 10, where 1 is the lowest priority and 10 is the highest. This
value should be left at the default.
- proxy_creds_lifetime
The amount of time (in minutes) that
proxy logins are valid. When a user first accesses a Remote Manager
process using an ACMS proxy, a login is created for the user, and the
expiration of that login is calculated and stored based on this
parameter. When the current time is greater than the expiration time,
the user's proxy information is refreshed. A change to this parameter
takes effect for any login that takes place after the change is made. A
change to this parameter does not take effect for any login that took
place before the change was made.
- rpc_priority
Relative priority of the RPC management thread.
The RPC management thread responds to RPC requests to get or set data
values. Priority is specified as a whole number between 1 and 10, where
1 is the lowest priority and 10 is the highest.
- snmp_agent_time_out
Number of seconds that the SNMP Master
agent waits for a response from the Remote Manager. The maximum is 10
seconds for Compaq TCP/IP Services Version 4.2. For Compaq TCP/IP
Services Version 5.0 and higher, the maximum is 60 seconds.
- snmp_are_you_there
Controls how often are you there messages
are sent by the Remote Manager to the SNMP Master agent. This value
should be entered as a multiple of the snmp_sel_time_out value. Each
time a timeout occurs, a timeout counter is incremented. The product of
the timeout counter and the snmp_sel_timeout are then compared to the
snmp_are_you_there value. If the product is greater than the
snmp_are_you_there value, an are_you_there message is sent.
- snmp_priority
Relative priority of the SNMP management thread.
The SNMP management thread responds to SNMP requests to get or set data
values. Priority is specified as a whole number between 1 and 10, where
1 is the lowest and 10 is the highest.
- snmp_sel_time_out
Controls how long the Remote Manager waits
for a response from the SNMP master agent. If the timeout valueis
reached and no messages are expected, the snmp_are_you_there interval
is checked (see snmp_are_you_there) . If a message is expected and is
not received before the select times out, the connection to the master
agent is assumed to have been lost and an attempt is made to
reregister. There is a hard coded 2 second wait prior to
reregistration.
This value also controls how long it takes to begin
disabling this interface. Requests to disable the interface do not
interrupt the socket select---they wait for it to either timeout or end
naturally (that is, when a message is received). At worse case, a
request to disable the interface has to wait snmp_sel_time_out seconds
before the shutdown of the interface begins. Once it begins, it usually
shuts down quickly---within a second or two.
- timer_interval
The Remote Manager runs one internal timer that
controls the operation of all other timers. The interval of this timer
effectively sets the smallest timer interval for the process. The
interval is set in seconds. If the value is too small, the timer will
run frequently with no work to do. This value should be set to smallest
desired timer interval.
- timer_priority
Relative priority of the timer thread. The timer
thread manages all internal timers. Priority is specified as a whole
number between 1 and 10, where 1 is the lowest priority and 10 is the
highest. This value should be left at the default.
- total_entity_slots
The total number of Collection table entries
to allow. When this number is reached, additional ACMSMGR ADD
COLLECTION requests are rejected. Slots are allocated when the ACMS
run-time system is started.
- trace_msg_wait_time
The number of seconds the Remote Manager
should wait for updates to the mss_coll_interval and wksp_coll_interval
parameters to become effective (processed by the ACC). Updates to the
ACC are sent by means of the trace monitor. The Remote Manager will
poll the value being changed for up to trace_msg_wait_time seconds to
see whether the value was in fact changed. If it is not changed within
this timeframe, the Remote Manager logs an error and returns an error
to the caller.
- trace_start_wait_time
The number of seconds the Remote Manager
should wait for the trace monitor to be started. The Remote Manager
communicates to ACMS process through the trace monitor. The Remote
Manager attempts to start the trace monitor if the Remote Manager needs
to send a message and the trace monitor is not already running. This
value controls how long the Remote Manager will wait for the trace
monitor to start before aborting the message send. Messages that are
not sent are discarded (lost).
- trap_priority
Relative priority of the trap sender thread. The
trap sender thread dispatches trap messages to SNMP and RPC receivers.
Priority is specified as a whole number between 1 and 10, where 1 is
the lowest priority and 10 is the highest. This value should be left at
the default.
- wksp_coll_interval
Controls the frequency (in seconds) at
which workspace (WS, WSC, TWS, TWSC) pool values are collected. A lower
value causes workspace values to be collected more often; a higher
value causes workspace pool values to be collected less often.
Workspace pool values are collected only by ACC and EXC.