svcadm(8)을 검색하려면 섹션에서 8 을 선택하고, 맨 페이지 이름에 svcadm을 입력하고 검색을 누른다.
nscd.conf(5)
nscd.conf(5) File Formats nscd.conf(5)
NAME
nscd.conf - name service cache daemon configuration
SYNOPSIS
svc:/system/name-service/cache
/etc/nscd.conf
DESCRIPTION
The SMF service svc:/system/name-service/cache contains the configura‐
tion information for nscd(8).
Do not edit the /etc/nscd.conf file as it is private.
Service configuration is applied from built-in defaults or from the
config property group. Provide Individual Cache configuration proper‐
ties from within their specific property group, or when unspecified by
the default property group.
Service Configuration Properties
The following table shows the service configuration properties that are
in the config property group:
tab(); lw(2.75i) lw(2.75i) lw(2.75i) lw(2.75i) PROPERTYUSAGE _ con‐
fig/enable_per_user_lookupEnable per-user lookups con‐
fig/per_user_nscd_time_to_livePer-user nscd time to live config/log‐
fileLog file location config/debug_levelDebugging/Log level flags con‐
fig/debug_componentsSubsystem debugging con‐
fig/global_check_smf_state_intervalSMF status check interval
enable_per_user_lookup value
Explicitly enable or disable nscd per-user mode, see nscd(8). By
default, the value is unset in SMF, which causes nscd to determine
whether the following conditions are true at startup:
o ldap is configured in nsswitch.conf(5)
o svc:/network/ldap/client (ldap_cachemgr(8)) is enabled
o ldapclient(8) has credentialLevel set to self.
Valid values for this property are true and false.
When enabled, multiple nscd processes exist. Two nscd processes
that have a UID of 0 (the parent and a forker) are always running,
while other per-user nscd processes that have a non-0 UID only run
as needed.
Using per-user lookups might not be possible if the corresponding
name service switch backends do not support those lookups or if the
backends are not configured to do so. Currently, ldap(7) is the
only applicable backend, see ldapclient(8) for information about
the credentialLevel and the authenticationMethod properties.
per_user_nscd_time_to_live value
Sets the idle time-to-live value for the non-0 UID per-user nscd
processes. The value, in seconds, is based on the last time the
per-user nscd was active. The default value is 120 seconds.
logfile filename
Specifies name of the file in which to write debug information. By
default, the file is /dev/null. Logging depends on the debug_level
and debug_components settings. Note that the log file can grow to
multiple gigabytes very quickly.
global_check_smf_state_interval value
Interval between checking the status of dependent SMF services such
as ldap(7), nis(7) and avahi-daemon(8). The default value is 120
seconds.
debug_level value
Specifies the desired debug level. The default value is "None". Use
the svccfg -s name-service/cache describe -tv config/debug_level
command to list available levels. Note that the configuration val‐
ues and the debug level output can change at any time, so do not
depend upon this information in scripts.
debug_components value
Specifies the nscd components that should generate debug messages.
The default value is "None". Use the
svccfg -s name-service/cache describe -tv config/debug_components
command to view the available components and a brief description of
each. Note that the configuration values and the component output
can change at any time, so do not depend upon this information in
scripts.
Cache Configuration Groups
You can configure a specific cache in its individual property group.
The following table summarizes the supported databases, their respec‐
tive SMF service property group names, and the entities that use them.
The Database column shows the form used by nscd usage output, while the
Group column shows the name used for the SMF property group.
tab(); lw(1.03i) lw(1.03i) lw(3.44i) lw(1.03i) lw(1.03i) lw(3.44i)
DatabaseGroupUsed By _ auth_attrauth_attrgetauthnam(3C) automountauto‐
mountautomount(8)bootparamsbootparamrpc.bootparamd(8) ether‐
setherethers(3C)exec_attrexec_attrexec_attr(5)groupgroupgetgrnam(3C)
hostshostT{ getaddrinfo(3C), getnameinfo(3C), gethostent(3C). T} ipn‐
odeshostT{ getaddrinfo(3C), getnameinfo(3C), getipnodebyaddr(3C). T}
netgroupnetgroupinnetgr(3C)netmasksnetmaskipadm(8) networksnetworkget‐
netbyname(3C)passwdpasswordgetpwnam(3C), getspnam(3C) prof_attr‐
prof_attrgetprofnam(3C), getexecprof(3C) projectprojectT{ getpro‐
jent(3PROJECT), getdefaultproj(3PROJECT), inproj(3PROJECT), newtask(1),
setproject(3PROJECT) T} protocolsprotocolgetprotobyname(3C) rpcrpcgetr‐
pcbyname(3C)servicesservicegetservbyname(3C)tnrhtptnrhtptnctl(8)tnrhdbtnrhdbtnctl(8)user_attruser_attrgetuserattr(3C)
Note that the ipnodes cache uses host group properties while being
cached independently.
Cache Configuration Properties
The following cache configuration properties are available within a
Group or the default property group:
check_file_interval value
Specifies an interval, in seconds, for the minimum time between
check_files checks. The default value is 0, which performs a check
each time the associated database is queried.
check_files value
Enables or disables the periodic checking of the corresponding
cache file. The time period is based on the number of seconds spec‐
ified by the check_file_interval property. By default, the periodic
check is enabled. The checks include the file's modified date,
size, and inode number. In addition, the check invalidates the
cache if any of the checked values differ from when the file was
last read. If you disable these checks you might experience a
slight performance boost. Disable these checks when you do not mod‐
ify the file and run the nscd -i cachename command as needed. Valid
values are true and false.
enable_cache value
Enables or disables the specified cache. The value can be either
true or false.
keep_hot_count value
Use this attribute to specify the number of entries that nscd(8)
keeps current in the specified cache. The value is an integer that
approximates the number of entries that are used frequently during
the day.
negative_time_to_live value
Specifies the time-to-live, in seconds, for unsuccessful queries
(negative entries) of the specified cache. Keep this value small to
reduce cache coherency issues.
positive_time_to_live value
Specifies the time-to-live, in seconds, for successful queries
(positive entries) of the specified cache. Larger values increase
cache hit rates and reduce mean response times and might increase
problems with cache coherence. Set the value to at least 43200 sec‐
onds (12 hours) to improve performance for sites that update NIS
maps nightly. When DNS is the source, use the DNS-returned TTL
value unless it is greater than the positive_time_to_live property
value. See nsswitch.conf(5).
Restart nscd after any SMF service changes
The nscd process must be restarted for any SMF service changes to take
effect.
example# svcadm restart name-service/cache
EXAMPLES
Example 1 Setting the passwd Cache's negative_time_to_live Property
Value to Ten Seconds.
Specifying a larger value might result in significant performance im‐
provements if several files are owned by user IDs (UIDs) that are not
in system databases.
example# svccfg -s name-service/cache
svc:/system/name-service/cache> addpg password application
svc:/system/name-service/cache> setprop password/negative_time_to_live = 10
svc:/system/name-service/cache> refresh
Note that you must add a property group before you can set the group's
properties.
Example 2 Disable the host and ipnode Cache
Disabling the host (or any other) cache may negatively impact perfor‐
mance but is sometimes desired when using certain third party products.
example# svccfg -s name-service/cache
svc:/system/name-service/cache> addpg host application
svc:/system/name-service/cache> setprop host/enable_cache = false
svc:/system/name-service/cache> refresh
Note that you must add a property group before you can set the proper‐
ties to that group.
Example 3 Re-enable a disabled host Cache
Re-enabling a previously disabled host (or any other) cache is straight
forward.
example# svccfg -s name-service/cache
svc:/system/name-service/cache> delcust host/enable_cache
svc:/system/name-service/cache> refresh
By default all caches are enabled, so this step is simply reversing a
previous administrator operation.
Example 4 Record minimal nscd diagnostics
Record minimal nscd diagnostics of incoming requests from processes and
report the success or failure of the requested lookups. Diagnostics
will be written to /usr/tmp/nscdlog. Setting the values in SMF causes
the diagnostics to be written indefinitely across reboots until dis‐
abled.
example# svccfg -s name-service/cache
svc:/system/name-service/cache> setprop config/logfile = "/usr/tmp/nscdlog"
svc:/system/name-service/cache> setprop config/debug_level = "all"
svc:/system/name-service/cache> setprop config/debug_components = "(" "frontend" "cache" ")"
svc:/system/name-service/cache> refresh
Note this can be useful when attempting to track down which process is
performing unexpected or undesired lookups. See nscd(8) to enable diag‐
nostics only on the active nscd process using nscd command line options
on a live system.
example# /usr/sbin/nscd -D all -C frontend,cache -L /usr/tmp/nscdlog
Example 5 Record full nscd diagnostics
Record full nscd diagnostics of all activity and details of all incom‐
ing requests for all lookups. Diagnostics will be written to the file
/usr/tmp/nscdlog. Setting the values in SMF causes the diagnostics to
be written indefinitely across reboots until disabled. Depending on how
busy a system is, this amount of diagnostics can generate gigabytes
(GB) of output per second so caution should be used.
example# svccfg -s name-service/cache
svc:/system/name-service/cache> setprop config/logfile = "/usr/tmp/nscdlog"
svc:/system/name-service/cache> setprop config/debug_level = "all"
svc:/system/name-service/cache> setprop config/debug_components = "all"
svc:/system/name-service/cache> refresh
Note this may be useful in recording early boot naming problems. See
nscd(8) to enable diagnostics only on the active nscd process using
nscd command line options on a live system.
example# /usr/sbin/nscd -D all -C all -L /usr/tmp/nscdlog
Example 6 Disable nscd diagnostics
Disable previously enabled diagnostics by removing previously set prop‐
erties.
example# svccfg -s name-service/cache
svc:/system/name-service/cache> delcust config/logfile
svc:/system/name-service/cache> delcust config/debug_level
svc:/system/name-service/cache> delcust config/debug_components
svc:/system/name-service/cache> refresh
Use this to turn off administrator enabled nscd diagnostics. See
nscd(8) to disable diagnostics only on the active nscd process using
nscd command line options on a live system.
example# /usr/sbin/nscd -D none -C none -L /dev/null
ATTRIBUTES
See attributes(7) for descriptions of the following attributes:
tab() box; cw(2.75i) |cw(2.75i) lw(2.75i) |lw(2.75i) ATTRIBUTE TYPEAT‐
TRIBUTE VALUE _ Availabilitysystem/core-os _ Interface StabilityCommit‐
ted
SEE ALSO
auth_attr(5), bootparams(5), ethers(5), exec_attr(5), group(5),
hosts(5), netmasks(5), networks(5), nsswitch.conf(5), passwd(5),
prof_attr(5), project(5), protocols(5), rpc(5), services(5),
user_attr(5), attributes(7), ldap(7), nis(7), avahi-daemon(8),
ldap_cachemgr(8), nscd(8), svccfg(8)
HISTORY
The Solaris 2.5 OS introduced the /etc/nscd.conf file as the nscd ad‐
ministrative configuration file.
The Oracle Solaris 10 OS introduced the svc:/system/name-service-cache
service. Having this service enabled has been recommended since Solaris
10 8/07.
The Oracle Solaris 11.0 OS introduced the svc:/system/name-ser‐
vice/cache service and replaced the /etc/nscd.conf administrative con‐
figuration file. Starting with Oracle Solaris 11.4, the service must be
enabled and be online with the nscd daemon running for certain configu‐
rations to function correctly. As of 11.4.63 nscd must be running for
lookups to any database backends other than files or DNS.
Starting with Oracle Solaris 11.4.57, the obsolete svc:/system/name-
service-cache was removed.
Oracle Solaris 11.4 29 Jul 2024 nscd.conf(5)