Chapter 3. Core Server Configuration Reference

This attribute specifies the path and filename of the log used to record each LDAP access. The following information is recorded by default in the log file:

For more information on turning access logging off, see the "Monitoring Server and Database Activity" chapter in the Red Hat Directory Server Administration Guide.

For access logging to be enabled, this attribute must have a valid path and parameter, and the nsslapd-accesslog-logging-enabled configuration attribute must be switched to on. The table lists the four possible combinations of values for these two configuration attributes and their outcome in terms of disabling or enabling of access logging.

This attribute controls what is logged to the access log.

You do not have to restart the server for this setting to take effect.

This read-only attribute, which cannot be set, provides a list of access log files used in access log rotation.

When set to off, the server writes all access log entries directly to disk. Buffering allows the server to use access logging even when under a heavy load without impacting performance. However, when debugging, it is sometimes useful to disable buffering in order to see the operations and their results right away instead of having to wait for the log entries to be flushed to the file. Disabling log buffering can severely impact performance in heavily loaded servers.

This attribute specifies the maximum age that a log file is allowed to reach before it is deleted. This attribute supplies only the number of units. The units are provided by the nsslapd-accesslog-logexpirationtimeunit attribute.

This attribute specifies the units for nsslapd-accesslog-logexpirationtime attribute. If the unit is unknown by the server, then the log never expires.

Disables and enables accesslog logging but only in conjunction with the nsslapd-accesslog attribute that specifies the path and parameter of the log used to record each database access.

For access logging to be enabled, this attribute must be switched to on, and the nsslapd-accesslog configuration attribute must have a valid path and parameter. The table lists the four possible combinations of values for these two configuration attributes and their outcome in terms of disabling or enabling of access logging.

This attribute specifies the maximum amount of disk space in megabytes that the access logs are allowed to consume. If this value is exceeded, the oldest access log is deleted.

When setting a maximum disk space, consider the total number of log files that can be created due to log file rotation. Also, remember that there are three different log files (access log, audit log, and error log) maintained by the Directory Server, each of which consumes disk space. Compare these considerations to the total amount of disk space for the access log.

This attribute sets the minimum allowed free disk space in megabytes. When the amount of free disk space falls below the value specified on this attribute, the oldest access logs are deleted until enough disk space is freed to satisfy this attribute.

This attribute sets whether access log rotation is to be synchronized with a particular time of the day. Synchronizing log rotation this way can generate log files at a specified time during a day, such as midnight to midnight every day. This makes analysis of the log files much easier because they then map directly to the calendar.

For access log rotation to be synchronized with time-of-day, this attribute must be enabled with the nsslapd-accesslog-logrotationsynchour and nsslapd-accesslog-logrotationsyncmin attribute values set to the hour and minute of the day for rotating log files.

For example, to rotate access log files every day at midnight, enable this attribute by setting its value to on, and then set the values of the nsslapd-accesslog-logrotationsynchour and nsslapd-accesslog-logrotationsyncmin attributes to 0.

This attribute sets the hour of the day for rotating access logs. This attribute must be used in conjunction with nsslapd-accesslog-logrotationsync-enabled and nsslapd-accesslog-logrotationsyncmin attributes.

This attribute sets the minute of the day for rotating access logs. This attribute must be used in conjunction with nsslapd-accesslog-logrotationsync-enabled and nsslapd-accesslog-logrotationsynchour attributes.

This attribute sets the time between access log file rotations. This attribute supplies only the number of units. The units (day, week, month, and so forth) are given by the nsslapd-accesslog-logrotationtimeunit attribute.

Directory Server rotates the log at the first write operation after the configured interval has expired, regardless of the size of the log.

Although it is not recommended for performance reasons to specify no log rotation since the log grows indefinitely, there are two ways of specifying this. Either set the nsslapd-accesslog-maxlogsperdir attribute value to 1 or set the nsslapd-accesslog-logrotationtime attribute to -1. The server checks the nsslapd-accesslog-maxlogsperdir attribute first, and, if this attribute value is larger than 1, the server then checks the nsslapd-accesslog-logrotationtime attribute. See Section 3.1.1.16, “nsslapd-accesslog-maxlogsperdir (Access Log Maximum Number of Log Files)” for more information.

This attribute sets the units for the nsslapd-accesslog-logrotationtime attribute.

This attribute sets the maximum access log size in megabytes. When this value is reached, the access log is rotated. That means the server starts writing log information to a new log file. If the nsslapd-accesslog-maxlogsperdir attribute is set to 1, the server ignores this attribute.

When setting a maximum log size, consider the total number of log files that can be created due to log file rotation. Also, remember that there are three different log files (access log, audit log, and error log) maintained by the Directory Server, each of which consumes disk space. Compare these considerations to the total amount of disk space for the access log.

This attribute sets the total number of access logs that can be contained in the directory where the access log is stored. Each time the access log is rotated, a new log file is created. When the number of files contained in the access log directory exceeds the value stored in this attribute, then the oldest version of the log file is deleted. For performance reasons, Red Hat recommends not setting this value to 1 because the server does not rotate the log, and it grows indefinitely.

If the value for this attribute is higher than 1, then check the nsslapd-accesslog-logrotationtime attribute to establish whether log rotation is specified. If the nsslapd-accesslog-logrotationtime attribute has a value of -1, then there is no log rotation. See Section 3.1.1.13, “nsslapd-accesslog-logrotationtime (Access Log Rotation Time)” for more information.

Note that, depending on the values set in nsslapd-accesslog-logminfreediskspace and nsslapd-accesslog-maxlogsize, the actual number of logs could be less than what you configure in nsslapd-accesslog-maxlogsperdir. For example, if nsslapd-accesslog-maxlogsperdir uses the default (10 files) and you set nsslapd-accesslog-logminfreediskspace to 500 MB and nsslapd-accesslog-maxlogsize to 100 MB, Directory Server keeps only 5 access files.

This attribute sets the access mode or file permission with which access log files are to be created. The valid values are any combination of 000 to 777 (these mirror the numbered or absolute UNIX file permissions). The value must be a 3-digit number, the digits varying from 0 through 7:

In the 3-digit number, the first digit represents the owner’s permissions, the second digit represents the group’s permissions, and the third digit represents everyone’s permissions. When changing the default value, remember that 000 does not allow access to the logs and that allowing write permissions to everyone can result in the logs being overwritten or deleted by anyone.

The newly configured access mode only affects new logs that are created; the mode is set when the log rotates to a new file.

If a user attempts to connect to the Directory Server without supplying any bind DN or password, this is an anonymous bind. Anonymous binds simplify common search and read operations, like checking the directory for a phone number or email address, by not requiring users to authenticate to the directory first.

However, there are risks with anonymous binds. Adequate ACIs must be in place to restrict access to sensitive information and to disallow actions like modifies and deletes. Additionally, anonymous binds can be used for denial of service attacks or for malicious people to gain access to the server.

Anonymous binds can be disabled to increase security (off). By default, anonymous binds are allowed (on) for search and read operations. This allows access to regular directory entries, which includes user and group entries as well as configuration entries like the root DSE. A third option, rootdse, allows anonymous search and read access to search the root DSE itself, but restricts access to all other directory entries.

Optionally, resource limits can be placed on anonymous binds using the nsslapd-anonlimitsdn attribute as described in Section 3.1.1.22, “nsslapd-anonlimitsdn”.

Changes to this value will not take effect until the server is restarted.

This parameter disables the pre-hashed password checks. By default, the Directory Server does not allow pre-hashed passwords to be set by anyone other than the Directory Manager. You can delegate this privilege to other users when you add them to the Password Administrators group. However in some scenarios, like when the replication partner already controls the pre-hashed passwords checking, this feature has to be disabled on the Directory Server.

Unauthenticated binds are connections to Directory Server where a user supplies an empty password. Using the default settings, Directory Server denies access in this scenario for security reasons.

You do not have to restart the server for this setting to take effect.

Per default, the root DSE lists all mechanisms the SASL library supports. However in some environments only certain ones are preferred. The nsslapd-allowed-sasl-mechanisms attribute allows you to enable only some defined SASL mechanisms.

The mechanism names must consist of uppercase letters, numbers, and underscores. Each mechanism can be separated by commas or spaces.

This setting does not require a server restart to take effect.

Resource limits can be set on authenticated binds. The resource limits can set a cap on how many entries can be searched in a single operation (nsslapd-sizeLimit), a time limit (nsslapd-timelimit) and time out period (nsslapd-idletimeout) for searches, and the total number of entries that can be searched (nsslapd-lookthroughlimit). These resource limits prevent denial of service attacks from tying up directory resources and improve overall performance.

Resource limits are set on a user entry. An anonymous bind, obviously, does not have a user entry associated with it. This means that resource limits usually do not apply to anonymous operations.

To set resource limits for anonymous binds, a template entry can be created, with the appropriate resource limits. The nsslapd-anonlimitsdn configuration attribute can then be added that points to this entry and applies the resource limits to anonymous binds.

This attribute allows non-standard characters in attribute names to be used for backwards compatibility with older servers, such as "_" in schema-defined attributes.

This attribute sets the path and filename of the log used to record changes made to each database.

For audit logging to be enabled, this attribute must have a valid path and parameter, and the nsslapd-auditlog-logging-enabled configuration attribute must be switched to on. The table lists the four possible combinations of values for these two configuration attributes and their outcome in terms of disabling or enabling of audit logging.

Provides a list of audit log files.

This attribute sets the maximum age that a log file is allowed to be before it is deleted. This attribute supplies only the number of units. The units (day, week, month, and so forth) are given by the nsslapd-auditlog-logexpirationtimeunit attribute.

This attribute sets the units for the nsslapd-auditlog-logexpirationtime attribute. If the unit is unknown by the server, then the log never expires.

Turns audit logging on and off.

For audit logging to be enabled, this attribute must have a valid path and parameter and the nsslapd-auditlog-logging-enabled configuration attribute must be switched to on. The table lists the four possible combinations of values for these two configuration attributes and their outcome in terms of disabling or enabling of audit logging.

This attribute sets the maximum amount of disk space in megabytes that the audit logs are allowed to consume. If this value is exceeded, the oldest audit log is deleted.

When setting a maximum disk space, consider the total number of log files that can be created due to log file rotation. Also remember that there are three different log files (access log, audit log, and error log) maintained by the Directory Server, each of which consumes disk space. Compare these considerations with the total amount of disk space for the audit log.

This attribute sets the minimum permissible free disk space in megabytes. When the amount of free disk space falls below the value specified by this attribute, the oldest audit logs are deleted until enough disk space is freed to satisfy this attribute.

This attribute sets whether audit log rotation is to be synchronized with a particular time of the day. Synchronizing log rotation this way can generate log files at a specified time during a day, such as midnight to midnight every day. This makes analysis of the log files much easier because they then map directly to the calendar.

For audit log rotation to be synchronized with time-of-day, this attribute must be enabled with the nsslapd-auditlog-logrotationsynchour and nsslapd-auditlog-logrotationsyncmin attribute values set to the hour and minute of the day for rotating log files.

For example, to rotate audit log files every day at midnight, enable this attribute by setting its value to on, and then set the values of the nsslapd-auditlog-logrotationsynchour and nsslapd-auditlog-logrotationsyncmin attributes to 0.

This attribute sets the hour of the day for rotating audit logs. This attribute must be used in conjunction with nsslapd-auditlog-logrotationsync-enabled and nsslapd-auditlog-logrotationsyncmin attributes.

This attribute sets the minute of the day for rotating audit logs. This attribute must be used in conjunction with nsslapd-auditlog-logrotationsync-enabled and nsslapd-auditlog-logrotationsynchour attributes.

This attribute sets the time between audit log file rotations. This attribute supplies only the number of units. The units (day, week, month, and so forth) are given by the nsslapd-auditlog-logrotationtimeunit attribute. If the nsslapd-auditlog-maxlogsperdir attribute is set to 1, the server ignores this attribute.

Directory Server rotates the log at the first write operation after the configured interval has expired, regardless of the size of the log.

Although it is not recommended for performance reasons to specify no log rotation, as the log grows indefinitely, there are two ways of specifying this. Either set the nsslapd-auditlog-maxlogsperdir attribute value to 1 or set the nsslapd-auditlog-logrotationtime attribute to -1. The server checks the nsslapd-auditlog-maxlogsperdir attribute first, and, if this attribute value is larger than 1, the server then checks the nsslapd-auditlog-logrotationtime attribute. See Section 3.1.1.37, “nsslapd-auditlog-maxlogsperdir (Audit Log Maximum Number of Log Files)” for more information.

This attribute sets the units for the nsslapd-auditlog-logrotationtime attribute.

This attribute sets the maximum audit log size in megabytes. When this value is reached, the audit log is rotated. That means the server starts writing log information to a new log file. If nsslapd-auditlog-maxlogsperdir to 1, the server ignores this attribute.

When setting a maximum log size, consider the total number of log files that can be created due to log file rotation. Also, remember that there are three different log files (access log, audit log, and error log) maintained by the Directory Server, each of which consumes disk space. Compare these considerations to the total amount of disk space for the audit log.

This attribute sets the total number of audit logs that can be contained in the directory where the audit log is stored. Each time the audit log is rotated, a new log file is created. When the number of files contained in the audit log directory exceeds the value stored on this attribute, then the oldest version of the log file is deleted. The default is 1 log. If this default is accepted, the server will not rotate the log, and it grows indefinitely.

If the value for this attribute is higher than 1, then check the nsslapd-auditlog-logrotationtime attribute to establish whether log rotation is specified. If the nsslapd-auditlog-logrotationtime attribute has a value of -1, then there is no log rotation. See Section 3.1.1.34, “nsslapd-auditlog-logrotationtime (Audit Log Rotation Time)” for more information.

This attribute sets the access mode or file permissions with which audit log files are to be created. The valid values are any combination of 000 to 777 since they mirror numbered or absolute UNIX file permissions. The value must be a combination of a 3-digit number, the digits varying from 0 through 7:

In the 3-digit number, the first digit represents the owner’s permissions, the second digit represents the group’s permissions, and the third digit represents everyone’s permissions. When changing the default value, remember that 000 does not allow access to the logs and that allowing write permissions to everyone can result in the logs being overwritten or deleted by anyone.

The newly configured access mode only affects new logs that are created; the mode is set when the log rotates to a new file.

This attribute sets the path and filename of the log used to record failed LDAP modifications.

If nsslapd-auditfaillog-logging-enabled is enabled, and nsslapd-auditfaillog is not set, the audit fail events are logged to the file specified in nsslapd-auditlog.

If you set the nsslapd-auditfaillog parameter to the same path as nsslapd-auditlog, both are logged in the same file.

To enable the audit fail log, this attribute must have a valid path and the nsslapd-auditfaillog-logging-enabled attribute must be set to on

Provides a list of audit fail log files.

This attribute sets the maximum age of a log file before it is removed. It supplies to the number of units. Specify the units, such as day, week, month, and so forth in the nsslapd-auditfaillog-logexpirationtimeunit attribute.

This attribute sets the units for the nsslapd-auditfaillog-logexpirationtime attribute. If the unit is unknown by the server, the log never expires.

Turns on and off logging of failed LDAP modifications.

This attribute sets the maximum amount of disk space in megabytes the audit fail logs are can consume. If the size exceed the limit, the oldest audit fail log is deleted.

This attribute sets the minimum permissible free disk space in megabytes. When the amount of free disk space is lower than the specified value, the oldest audit fail logs are deleted until enough disk space is freed.

This attribute sets whether audit fail log rotation is to be synchronized with a particular time of the day. Synchronizing log rotation this way can generate log files at a specified time during a day, such as midnight to midnight every day. This makes analysis of the log files much easier because they then map directly to the calendar.

For audit fail log rotation to be synchronized with time-of-day, this attribute must be enabled with the nsslapd-auditfaillog-logrotationsynchour and nsslapd-auditfaillog-logrotationsyncmin attribute values set to the hour and minute of the day for rotating log files.

For example, to rotate audit fail log files every day at midnight, enable this attribute by setting its value to on, and then set the values of the nsslapd-auditfaillog-logrotationsynchour and nsslapd-auditfaillog-logrotationsyncmin attributes to 0.

This attribute sets the hour of the day the audit fail log is rotated. This attribute must be used in conjunction with nsslapd-auditfaillog-logrotationsync-enabled and nsslapd-auditfaillog-logrotationsyncmin attributes.

This attribute sets the minute the audit fail log is rotated. This attribute must be used in conjunction with nsslapd-auditfaillog-logrotationsync-enabled and nsslapd-auditfaillog-logrotationsynchour attributes.

This attribute sets the time between audit fail log file rotations. This attribute supplies only the number of units. The units (day, week, month, and so forth) are given by the nsslapd-auditfaillog-logrotationtimeunit attribute. If the nsslapd-auditfaillog-maxlogsperdir attribute is set to 1, the server ignores this attribute.

Directory Server rotates the log at the first write operation after the configured interval has expired, regardless of the size of the log.

Although it is not recommended for performance reasons to specify no log rotation, as the log grows indefinitely, there are two ways of specifying this. Either set the nsslapd-auditfaillog-maxlogsperdir attribute value to 1 or set the nsslapd-auditfaillog-logrotationtime attribute to -1. The server checks the nsslapd-auditfaillog-maxlogsperdir attribute first, and, if this attribute value is larger than 1, the server then checks the nsslapd-auditfaillog-logrotationtime attribute. See Section 3.1.1.52, “nsslapd-auditfaillog-maxlogsperdir (Audit Fail Log Maximum Number of Log Files)” for more information.

This attribute sets the units for the nsslapd-auditfaillog-logrotationtime attribute.

This attribute sets the maximum audit fail log size in megabytes. When this value is reached, the audit fail log is rotated. That means the server starts writing log information to a new log file. If the nsslapd-auditfaillog-maxlogsperdir parameter is set to 1, the server ignores this attribute.

This attribute sets the total number of audit fail logs that can be contained in the directory where the audit log is stored. Each time the audit fail log is rotated, a new log file is created. When the number of files contained in the audit log directory exceeds the value stored on this attribute, then the oldest version of the log file is deleted. The default is 1 log. If this default is accepted, the server will not rotate the log, and it grows indefinitely.

If the value for this attribute is higher than 1, then check the nsslapd-auditfaillog-logrotationtime attribute to establish whether log rotation is specified. If the nsslapd-auditfaillog-logrotationtime attribute has a value of -1, then there is no log rotation. See Section 3.1.1.49, “nsslapd-auditfaillog-logrotationtime (Audit Fail Log Rotation Time)” for more information.

This attribute sets the access mode or file permissions with which audit fail log files are to be created. The valid values are any combination of 000 to 777 since they mirror numbered or absolute UNIX file permissions. The value must be a combination of a 3-digit number, the digits varying from 0 through 7:

In the 3-digit number, the first digit represents the owner’s permissions, the second digit represents the group’s permissions, and the third digit represents everyone’s permissions. When changing the default value, remember that 000 does not allow access to the logs and that allowing write permissions to everyone can result in the logs being overwritten or deleted by anyone.

The newly configured access mode only affects new logs that are created; the mode is set when the log rotates to a new file.

This parameter sets the path to the default backup directory. The Directory Server user must have write permissions in the configured directory.

This setting does not require a server restart to take effect.

This parameter defines the full path to the directory that Directory Server uses to store the Network Security Services (NSS) database of the instance. This database contains the private keys and certificates of the instance.

As a fallback, Directory Server extracts the private key and certificates to this directory, if the server cannot extract them to the /tmp/ directory in a private name space. For details about private name spaces, see the PrivateTmp parameter description in the systemd.exec(5) man page.

The directory specified in nsslapd-certdir must be owned by the user ID of the server, and only this user ID must have read-write permissions in this directory. For security reasons, no other users should have permissions to read or write to this directory.

The service must be restarted for changes to this attribute to take effect.

This attribute can be used when client authentication is performed using TLS certificates in order to avoid limitations of the security subsystem certificate mapping, configured in the /etc/dirsrv/slapd-instance_name/certmap.conf file. Depending on the configuration in this file, the certificate mapping may be done using a directory subtree search based at the root DN. If the search is based at the root DN, then the nsslapd-certmap-basedn attribute may force the search to be based at some entry other than the root. The valid value for this attribute is the DN of the suffix or subtree to use for certificate mapping.

This read-only attribute is the config DN.

This parameter allows you to enable a DN inside a CN value.

The Directory Server DN normalizer follows RFC4514 and keeps a white space if the RDN attribute type is not based on the DN syntax. However the Directory Server’s configuration entry sometimes uses a cn attribute to store a DN value. For example in dn: cn="dc=A,dc=com", cn=mapping tree,cn=config, the cn should be normalized following the DN syntax.

If this configuration is required, enable the nsslapd-cn-uses-dn-syntax-in-dns parameter.

This attribute sets the connection buffering behavior. Possible values:

The value 2 provides a better performance if the client sends a large amount of data at once. This is, for example, the case for large add and modify operations, or when many asynchronous requests are received over a single connections like during a replication.

This option allows you to enable or disable the SASL NOCANON flag. Disabling avoids the Directory Server looking up DNS reverse entries for outgoing connections.

This attribute sets the connection table size, which determines the total number of connections supported by the server.

Increase the value of this attribute if Directory Server is refusing connections because it is out of connection slots. When this occurs, the Directory Server’s error log file records the message Not listening for new connections — too many fds open.

It may be necessary to increase the operating system limits for the number of open files and number of open files per process, and it may be necessary to increase the ulimit for the number of open files (ulimit -n) in the shell that starts Directory Server.

The size of the connection table is cap with nsslapd-maxdescriptor. See Section 3.1.1.117, “nsslapd-maxdescriptors (Maximum File Descriptors)” for more information.

The server has to be restarted for changes to this attribute to go into effect.

The nsslapd-counters attribute enables and disables Directory Server database and server performance counters.

There can be a performance impact by keeping track of the larger counters. Turning off 64-bit integers for counters can have a minimal improvement on performance, although it negatively affects long term statistics tracking.

This parameter is enabled by default. To disable counters, stop the Directory Server, edit the dse.ldif file directly, and restart the server.

This attribute sets whether change sequence numbers (CSNs), when available, are to be logged in the access log. By default, CSN logging is turned on.

This attribute gives the naming context, of all configured naming contexts, which clients should use by default as a search base. This value is copied over to the root DSE as the defaultNamingContext attribute, which allows clients to query the root DSE to obtain the context and then to initiate a search with the appropriate base.

This attribute enables a thread which runs every ten (10) seconds to check the available disk space on the disk or mount where the Directory Server database is running. If the available disk space drops below a configured threshold, then the server begins reducing logging levels, disabling access or audit logs, and deleting rotated logs. If that does not free enough available space, then the server shuts down gracefully (after a wanring and grace period).

Sets a grace period to wait before shutting down the server after it hits half of the disk space limit set in Section 3.1.1.69, “nsslapd-disk-monitoring-threshold”. This gives the administrator time to clean out the disk and prevent a shutdown.

Sets whether to shut down the server if the log directories pass the halfway point set in the disk space limit, Section 3.1.1.69, “nsslapd-disk-monitoring-threshold”.

If this is enabled, then logging is not disabled and rotated logs are not deleted as means of reducing disk usage by the server. The server simply goes toward a shutdown process.

If the free disk space reaches half of the value you set in the nsslapd-disk-monitoring-threshold parameter, Directory Server shuts down the instance after the grace period set in nsslapd-disk-monitoring-grace-period is reached. However, if the disk runs out of space before the instance is down, data can be corrupted. To prevent this problem, enable the nsslapd-disk-monitoring-readonly-on-threshold parameter, the Directory Server sets the instance to read-only mode when the threshold is reached.

The service must be restarted for changes to this attribute to take effect.

Sets the threshold, in bytes, to use to evaluate whether the server has enough available disk space. Once the space reaches half of this threshold, then the server begins a shut down process.

For example, if the threshold is 2MB (the default), then once the available disk space reaches 1MB, the server will begin to shut down.

By default, the threshold is evaluated backs on the disk space used by the configuration, transaction, and database directories for the Directory Server instance. If the Section 3.1.1.67, “nsslapd-disk-monitoring-logging-critical” attribute is enabled, then the log directory is included in the evaluation.

The Section 3.1.1.166, “nsslapd-syntaxcheck” attribute enables the server to verify that any new or modified attribute value matches the required syntax for that attribute.

However, the syntax rules for DNs have grown increasingly strict. Attempting to enforce DN syntax rules in RFC 4514 could break many servers using older syntax definitions. By default, then nsslapd-syntaxcheck validates DNs using RFC 1779 or RFC 2253.

The nsslapd-dn-validate-strict attribute explicitly enables strict syntax validation for DNs, according to section 3 in RFC 4514. If this attribute is set to off (the default), the server normalizes the value before checking it for syntax violations.

Makes the schema in cn=schema compatible with 4.x versions of Directory Server.

The Directory Server turbo mode is a feature that enables a worker thread to be dedicated to a connection and continuously read incoming operations from that connection. This can improve the performance on very active connections, and the feature is enabled by default.

Worker threads are processing the LDAP operation received by the server. The number of worker threads is defined in the nsslapd-threadnumber parameter. Every five seconds, each worker thread evaluates if the activity level of its current connection is one of the highest among all established connections. Directory Server measures the activity as the number of operations initiated since the last check, and switches a worker thread in turbo mode if the activity of the current connection is one of the highest.

If you encounter long execution times (etime value in log files) for bind operations, such as one second or longer, deactivating the turbo mode can improve the performance. However, in some cases, long bind times are a symptom of networking or hardware issues. In these situations, disabling the turbo mode does not result in improved performance.

During a simple bind, Directory Server has access to the plain text password due to the nature of bind operations. If the nsslapd-enable-upgrade-hash parameter is enabled and a user authenticates, Directory Server checks if the userPassword attribute of the user uses the hashing algorithm set in the passwordStorageScheme attribute. If the algorithm is different, the server hashes the plain text password with the algorithm from passwordStorageScheme and updates the value of the user’s userPassword attribute.

For example, if you import a user entry with a password that is hashed using a weak algorithm, the server automatically re-hashes the passwords on the first login of the user using the algorithm set in passwordStorageScheme, which is, by default, PBKDF2_SHA256.

This attribute is deprecated and will be removed in a future version of Directory Server.

This attribute controls whether quoting in the objectclass attributes contained in the cn=schema entry conforms to the quoting specified by Internet draft RFC 2252. By default, the Directory Server conforms to RFC 2252, which indicates that this value should not be quoted. Only very old clients need this value set to on, so leave it off.

Turning this attribute on or off does not affect Directory Server Console.

The nsslapd-entryusn-global parameter defines if the USN plug-in assigns unique update sequence numbers (USN) across all back end databases or to each database individually. For unique USNs across all back end databases, set this parameter to on.

For further details, see Section 6.8, “entryusn”.

You do not have to restart the server for this setting to take effect.

Entry update sequence numbers (USNs) are not preserved when entries are exported from one server and imported into another, including when initializing a database for replication. By default, the entry USNs for imported entries are set to zero.

It is possible to configure a different initial value for entry USNs using nsslapd-entryusn-import-initval. This sets a starting USN which is used for all imported entries.

There are two possible values for nsslapd-entryusn-import-initval:

This attribute sets the path and filename of the log used to record error messages generated by the Directory Server. These messages can describe error conditions, but more often they contain informative conditions, such as:

This log contains differing amounts of information depending on the current setting of the Log Level attribute. See Section 3.1.1.78, “nsslapd-errorlog-level (Error Log Level)” for more information.

For error logging to be enabled, this attribute must have a valid path and filename, and the nsslapd-errorlog-logging-enabled configuration attribute must be switched to on. The table lists the four possible combinations of values for these two configuration attributes and their outcome in terms of disabling or enabling of error logging.

This attribute sets the level of logging for the Directory Server. The log level is additive; that is, specifying a value of 3 includes both levels 1 and 2.

The default value for nsslapd-errorlog-level is 16384.

You do not have to restart the server for this setting to take effect.

This read-only attribute provides a list of error log files.

This attribute sets the maximum age that a log file is allowed to reach before it is deleted. This attribute supplies only the number of units. The units (day, week, month, and so forth) are given by the nsslapd-errorlog-logexpirationtimeunit attribute.

This attribute sets the units for the nsslapd-errorlog-logexpirationtime attribute. If the unit is unknown by the server, then the log never expires.

Turns error logging on and off.

This attribute sets the maximum amount of disk space in megabytes that the error logs are allowed to consume. If this value is exceeded, the oldest error log is deleted.

When setting a maximum disk space, consider the total number of log files that can be created due to log file rotation. Also, remember that there are three different log files (access log, audit log, and error log) maintained by the Directory Server, each of which consumes disk space. Compare these considerations to the total amount of disk space for the error log.

This attribute sets the minimum allowed free disk space in megabytes. When the amount of free disk space falls below the value specified on this attribute, the oldest error log is deleted until enough disk space is freed to satisfy this attribute.

This attribute sets whether error log rotation is to be synchronized with a particular time of the day. Synchronizing log rotation this way can generate log files at a specified time during a day, such as midnight to midnight every day. This makes analysis of the log files much easier because they then map directly to the calendar.

For error log rotation to be synchronized with time-of-day, this attribute must be enabled with the nsslapd-errorlog-logrotationsynchour and nsslapd-errorlog-logrotationsyncmin attribute values set to the hour and minute of the day for rotating log files.

For example, to rotate error log files every day at midnight, enable this attribute by setting its value to on, and then set the values of the nsslapd-errorlog-logrotationsynchour and nsslapd-errorlog-logrotationsyncmin attributes to 0.

This attribute sets the hour of the day for rotating error logs. This attribute must be used in conjunction with nsslapd-errorlog-logrotationsync-enabled and nsslapd-errorlog-logrotationsyncmin attributes.

This attribute sets the minute of the day for rotating error logs. This attribute must be used in conjunction with nsslapd-errorlog-logrotationsync-enabled and nsslapd-errorlog-logrotationsynchour attributes.

This attribute sets the time between error log file rotations. This attribute supplies only the number of units. The units (day, week, month, and so forth) are given by the nsslapd-errorlog-logrotationtimeunit (Error Log Rotation Time Unit) attribute.

Directory Server rotates the log at the first write operation after the configured interval has expired, regardless of the size of the log.

Although it is not recommended for performance reasons to specify no log rotation, as the log grows indefinitely, there are two ways of specifying this. Either set the nsslapd-errorlog-maxlogsperdir attribute value to 1 or set the nsslapd-errorlog-logrotationtime attribute to -1. The server checks the nsslapd-errorlog-maxlogsperdir attribute first, and, if this attribute value is larger than 1, the server then checks the nsslapd-errorlog-logrotationtime attribute. See Section 3.1.1.91, “nsslapd-errorlog-maxlogsperdir (Maximum Number of Error Log Files)” for more information.

This attribute sets the units for nsslapd-errorlog-logrotationtime (Error Log Rotation Time). If the unit is unknown by the server, then the log never expires.

This attribute sets the maximum error log size in megabytes. When this value is reached, the error log is rotated, and the server starts writing log information to a new log file. If nsslapd-errorlog-maxlogsperdir is set to 1, the server ignores this attribute.

When setting a maximum log size, consider the total number of log files that can be created due to log file rotation. Also, remember that there are three different log files (access log, audit log, and error log) maintained by the Directory Server, each of which consumes disk space. Compare these considerations to the total amount of disk space for the error log.

This attribute sets the total number of error logs that can be contained in the directory where the error log is stored. Each time the error log is rotated, a new log file is created. When the number of files contained in the error log directory exceeds the value stored on this attribute, then the oldest version of the log file is deleted. The default is 1 log. If this default is accepted, the server does not rotate the log, and it grows indefinitely.

If the value for this attribute is higher than 1, then check the nsslapd-errorlog-logrotationtime attribute to establish whether log rotation is specified. If the nsslapd-errorlog-logrotationtime attribute has a value of -1, then there is no log rotation. See Section 3.1.1.88, “nsslapd-errorlog-logrotationtime (Error Log Rotation Time)” for more information.

This attribute sets the access mode or file permissions with which error log files are to be created. The valid values are any combination of 000 to 777 since they mirror numbered or absolute UNIX file permissions. That is, the value must be a combination of a 3-digit number, the digits varying from 0 through 7:

In the 3-digit number, the first digit represents the owner’s permissions, the second digit represents the group’s permissions, and the third digit represents everyone’s permissions. When changing the default value, remember that 000 does not allow access to the logs and that allowing write permissions to everyone can result in the logs being overwritten or deleted by anyone.

The newly configured access mode only affects new logs that are created; the mode is set when the log rotates to a new file.

When establishing a TLS connection, a client sends its certificate first and then issues a BIND request using the SASL/EXTERNAL mechanism. Using SASL/EXTERNAL tells the Directory Server to use the credentials in the certificate for the TLS handshake. However, some clients do not use SASL/EXTERNAL when they send their BIND request, so the Directory Server processes the bind as a simple authentication request or an anonymouse request and the TLS connection fails.

The nsslapd-force-sasl-external attribute forces clients in certificate-based authentication to send the BIND request using the SASL/EXTERNAL method.

This attribute is deprecated, and documented here only for historical purposes.

The Access Control Plug-in does not use the value specified by the nsslapd-groupevalnestlevel attribute to set the number of levels of nesting that access control performs for group evaluation. Instead, the number of levels of nesting is hardcoded as 5.

This attribute sets the amount of time in seconds after which an idle LDAP client connection is closed by the server. A value of 0 means that the server never closes idle connections. This setting applies to all connections and all users. Idle timeout is enforced when the connection table is walked, when poll() does not return zero. Therefore, a server with a single connection never enforces the idle timeout.

Use the nsIdleTimeout operational attribute, which can be added to user entries, to override the value assigned to this attribute. For details, see the "Setting Resource Limits Based on the Bind DN" section in the Red Hat Directory Server Administration Guide.

This parameter allows to disable the virtual attribute lookup in a search entry.

If you do not require virtual attributes, you can disable virtual attribute lookups in search results to increase the speed of searches.

This attribute is deprecated. There are now separate configuration parameters for instance-specific paths, such as nsslapd-certdir and nsslapd-lockdir. See the documentation for the specific directory path that is set.

This attribute sets the amount of time in milliseconds after which the connection to a stalled LDAP client is closed. An LDAP client is considered to be stalled when it has not made any I/O progress for read or write operations.

This attribute sets whether the Directory Server maintains the creatorsName, createTimestamp, modifiersName, and modifyTimestamp operational attributes for newly created or updated entries.

You do not have to restart the server for this setting to take effect.

The nsslapd-ldapiautobind sets whether the server will allow users to autobind to Directory Server using LDAPI. Autobind maps the UID or GUID number of a system user to a Directory Server user, and automatically authenticates the user to Directory Server based on those credentials. The Directory Server connection occurs over UNIX socket.

Along with enabling autobind, configuring autobind requires configuring mapping entries. The nsslapd-ldapimaprootdn maps a root user on the system to the Directory Manager. The nsslapd-ldapimaptoentries maps regular users to Directory Server users, based on the parameters defined in the nsslapd-ldapiuidnumbertype, nsslapd-ldapigidnumbertype, and nsslapd-ldapientrysearchbase attributes.

Autobind can only be enabled if LDAPI is enabled, meaning the nsslapd-ldapilisten is on and the nsslapd-ldapifilepath attribute is set to an LDAPI socket.

With autobind, it is possible to map system users to Directory Server user entries, based on the system user’s UID and GUID numbers. This requires setting Directory Server parameters for which attribute to use for the UID number (nsslapd-ldapiuidnumbertype) and GUID number (nsslapd-ldapigidnumbertype) and setting the search base to use to search for matching user entries.

The nsslapd-ldapientrysearchbase gives the subtree to search for user entries to use for autobind.

LDAPI connects a user to an LDAP server over a UNIX socket rather than TCP. In order to configure LDAPI, the server must be configured to communicate over a UNIX socket. The UNIX socket to use is set in the nsslapd-ldapifilepath attribute.

Autobind can be used to authenticate system users to the server automatically and connect to the server using a UNIX socket. To map the system user to a Directory Server user for authentication, the system user’s UID and GUID numbers should be mapped to be a Directory Server attribute. The nsslapd-ldapigidnumbertype attribute points to the Directory Server attribute to map system GUIDs to user entries.

Users can only connect to the server with autobind if LDAPI is enabled (nsslapd-ldapilisten and nsslapd-ldapifilepath), autobind is enabled (nsslapd-ldapiautobind), and autobind mapping is enabled for regular users (nsslapd-ldapimaptoentries).

The nsslapd-ldapilisten enables LDAPI connections to the Directory Server. LDAPI allows users to connect to the Directory Server over a UNIX socket rather than a standard TCP port. Along with enabling LDAPI by setting nsslapd-ldapilisten to on, there must also be a UNIX socket set for LDAPI in the nsslapd-ldapifilepath attribute.

With autobind, a system user is mapped to a Directory Server user and then automatically authenticated to the Directory Server over a UNIX socket.

The root system user (the user with a UID of 0) is mapped to whatever Directory Server entry is specified in the nsslapd-ldapimaprootdn attribute.

With autobind, a system user is mapped to a Directory Server user and then automatically authenticated to the Directory Server over a UNIX socket. This mapping is automatic for root users, but it must be enabled for regular system users through the nsslapd-ldapimaptoentries attribute. Setting this attribute to on enables mapping for regular system users to Directory Server entries. If this attribute is not enabled, then only root users can use autobind to authenticate to the Directory Server, and all other users connect anonymously.

The mappings themselves are configured through the nsslapd-ldapiuidnumbertype and nsslapd-ldapigidnumbertype attributes, which map Directory Server attributes to the user’s UID and GUID numbers.

Users can only connect to the server with autobind if LDAPI is enabled (nsslapd-ldapilisten and nsslapd-ldapifilepath) and autobind is enabled (nsslapd-ldapiautobind).

Autobind can be used to authenticate system users to the server automatically and connect to the server using a UNIX socket. To map the system user to a Directory Server user for authentication, the system user’s UID and GUID numbers must be mapped to be a Directory Server attribute. The nsslapd-ldapiuidnumbertype attribute points to the Directory Server attribute to map system UIDs to user entries.

Users can only connect to the server with autobind if LDAPI is enabled (nsslapd-ldapilisten and nsslapd-ldapifilepath), autobind is enabled (nsslapd-ldapiautobind), and autobind mapping is enabled for regular users (nsslapd-ldapimaptoentries).

Directory Server exports files in LDAP Data Interchange Format (LDIF) format to the directory set in this parameter when using the db2ldif or db2ldif.pl. The directory must be owned by the Directory Server user and group. Only this user and group must have read and write access in this directory.

The service must be restarted for changes to this attribute to take effect.

This attribute sets the maximum of the socket connection backlog. The listen service sets the number of sockets available to receive incoming connections. The backlog setting sets a maximum length for how long the queue for the socket (sockfd) can grow before refusing connections.

This attribute allows multiple Directory Server instances to run on a multihomed machine (or makes it possible to limit listening to one interface of a multihomed machine). There can be multiple IP addresses associated with a single hos tname, and these IP addresses can be a mix of both IPv4 and IPv6. This parameter can be used to restrict the Directory Server instance to a single IP interface.

If a host name is given as the nsslapd-listenhost value, then the Directory Server responds to requests for every interface associated with the host name. If a single IP interface (either IPv4 or IPv6) is given as the nsslapd-listenhost value, Directory Server only responds to requests sent to that specific interface. Either an IPv4 or IPv6 address can be used.

The server has to be restarted for changes to this attribute to go into effect.

This attribute specifies the host machine on which the Directory Server runs. This attribute creates the referral URL that forms part of the MMR protocol. In a high-availability configuration with failover nodes, that referral should point to the virtual name of the cluster, not the local host name.

This attribute sets the user as whom the Directory Server runs. The group as which the user runs is derived from this attribute by examining the user’s primary group. Should the user change, then all of the instance-specific files and directories for this instance need to be changed to be owned by the new user, using a tool such as chown.

The value for the nsslapd-localuser is set initially when the server instance is configured.

This is the full path to the directory the server uses for lock files. The default value is /var/lock/dirsrv/slapd-instance. Changes to this value will not take effect until the server is restarted.

The nsslapd-localssf parameter sets the security strength factor (SSF) for LDAPI connections. Directory Server allows LDAPI connections only if the value set in nsslapd-localssf is greater or equal than the value set in the nsslapd-minssf parameter. Therefore, LDAPI connections meet the minimum SSF set in nsslapd-minssf.

You do not have to restart the server for this setting to take effect.

Controls whether logs will use high resolution timestamps with nanosecond precision, or standard resolution timestamps with one second precision. Enabled by default. Set this option to off to revert log timestamps back to one second precision.

This setting does not require restarting the server to take effect.

Defines the maximum size in bytes allowed for an incoming message. This limits the size of LDAP requests that can be handled by the Directory Server. Limiting the size of requests prevents some kinds of denial of service attacks.

The limit applies to the total size of the LDAP request. For example, if the request is to add an entry and if the entry in the request is larger than the configured value or the default, then the add request is denied. However, the limit is not applied to replication processes. Be cautious before changing this attribute.

This setting does not require a server restart to take effect.

This attribute sets the maximum, platform-dependent number of file descriptors that the Directory Server tries to use. A file descriptor is used whenever a client connects to the server. File descriptors are also used by access logs, error logs, audit logs, database files (indexes and transaction logs), and as sockets for outgoing connections to other servers for replication and chaining.

The number of descriptors available for TCP/IP to serve client connections is determined by the nsslapd-conntablesize attribute. The default value for this attribute is set to the file descriptor soft limit, which defaults to 1024. However, if you configure this attribute manually, the server updates the process file descriptor soft limit to match.

If this value is set too high, the Directory Server queries the operating system for the maximum allowable value, and then uses that value. It also issues an information message in the error log. If this value is set to an invalid value remotely, by using the Directory Server Console or ldapmodify, the server rejects the new value, keeps the old value, and responds with an error.

Some operating systems let users configure the number of file descriptors available to a process. See the operating system documentation for details on file descriptor limits and configuration. The dsktune program (explained in the Red Hat Directory Server Installation Guide) can be used to suggest changes to the system kernel or TCP/IP tuning attributes, including increasing the number of file descriptors if necessary. Increased the value on this attribute if the Directory Server is refusing connections because it is out of file descriptors. When this occurs, the following message is written to the Directory Server’s error log file:

Not listening for new connections -- too many fds open

See Section 3.1.1.61, “nsslapd-conntablesize” for more information about increasing the number of incoming connections.

The server has to be restarted for changes to this attribute to go into effect.

When a user is authenticated to the Directory Server over SASL GSS-API, the server must allocate a certain amount of memory to the client to perform LDAP operations, according to how much memory the client requests. It is possible for an attacker to send such a large packet size that it crashes the Directory Server or ties it up indefinitely as part of a denial of service attack.

The packet size which the Directory Server will allow for SASL clients can be limited using the nsslapd-maxsasliosize attribute. This attribute sets the maximum allowed SASL IO packet size that the server will accept.

When an incoming SASL IO packet is larger than the nsslapd-maxsasliosize limit, the server immediately disconnects the client and logs a message to the error log, so that an administrator can adjust the setting if necessary.

This attribute value is specified in bytes.

Defines the maximum number of threads that a connection should use. For normal operations where a client binds and only performs one or two operations before unbinding, use the default value. For situations where a client binds and simultaneously issues many requests, increase this value to allow each connection enough resources to perform all the operations. This attribute is not available from the server console.

A security strength factor is a relative measurement of how strong a connection is according to its key strength. The SSF determines how secure an TLS or SASL connection is. The nsslapd-minssf attribute sets a minimum SSF requirement for any connection to the server; any connection attempts that are weaker than the minimum SSF are rejected.

TLS and SASL connections can be mixed in a connection to the Directory Server. These connections generally have different SSFs. The higher of the two SSFs is used to compare to the minimum SSF requirement.

Setting the SSF value to 0 means that there is no minimum setting.

A security strength factor is a relative measurement of how strong a connection is according to its key strength. The SSF determines how secure an TLS or SASL connection is.

The nsslapd-minssf-exclude-rootdse attribute sets a minimum SSF requirement for any connection to the server except for queries for the root DSE. This enforces appropriate SSF values for most connections, while still allowing clients to get required information about the server configuration from the root DSE without having to establish a secure connection first.

This parameter controls the ACI checks when directory entries are moved from one subtree to another and using source and target restrictions in moddn operations. For backward compatibility, you can disable the ACI checks.

If a Directory Server instance is started as a service using the systemctl utility, environment variables are not passed to the server unless you set them in the /etc/sysconfig/dirsrv or /etc/sysconfig/dirsrv-instance_name file. For further details, see the systemd.exec(3) man page.

Instead of manually editing the service files to set the M_MMAP_THRESHOLD environment variable, the nsslapd-malloc-mmap-threshold parameter enables you to set the value in the Directory Server configuration. For further details, see the M_MMAP_THRESHOLD parameter description in the mallopt(3) man page.

This setting does not require restarting the server to take effect.

If a Directory Server instance is started as a service using the systemctl utility, environment variables are not passed to the server unless you set them in the /etc/sysconfig/dirsrv or /etc/sysconfig/dirsrv-instance_name file. For further details, see the systemd.exec(3) man page.

Instead of manually editing the service files to set the M_MXFAST environment variable, the nsslapd-malloc-mxfast parameter enables you to set the value in the Directory Server configuration. For further details, see the M_MXFAST parameter description in the mallopt(3) man page.

This setting does not require restarting the server to take effect.

If a Directory Server instance is started as a service using the systemctl utility, environment variables are not passed to the server unless you set them in the /etc/sysconfig/dirsrv or /etc/sysconfig/dirsrv-instance_name file. For further details, see the systemd.exec(3) man page.

Instead of manually editing the service files to set the M_TRIM_THRESHOLD environment variable, the nsslapd-malloc-trim-threshold parameter enables you to set the value in the Directory Server configuration. For further details, see the M_TRIM_THRESHOLD parameter description in the mallopt(3) man page.

This setting does not require restarting the server to take effect.

When the value of this attribute is off, the TCP_NODELAY option is set so that LDAP responses (such as entries or result messages) are sent back to a client immediately. When the attribute is turned on, default TCP behavior applies; specifically, sending data is delayed so that additional data can be grouped into one packet of the underlying network MTU size, typically 1500 bytes for Ethernet.

Normalizing distinguished names (DN) is a resource intensive task. If the nsslapd-ndn-cache-enabled parameter is enabled, Directory Server caches normalized DNs in memory. Update the nsslapd-ndn-cache-max-size parameter to set the maximum size of this cache.

Normalizing distinguished names (DN) is a resource intensive task. If the nsslapd-ndn-cache-enabled parameter is enabled, Directory Server caches normalized DNs in memory. The nsslapd-ndn-cache-max-size parameter sets the maximum size of this cache.

If a DN requested is not cached already, it is normalized and added. When the cache size limit is exceeded, Directory Server removes the least recently used 10,000 DNs from the cache. However, a minimum of 10,000 DNs is always kept cached.

This attribute limits the I/O wait time for all outbound LDAP connections. The default is 300000 milliseconds (5 minutes). A value of 0 means that the server does not impose a limit on I/O wait time.

This attribute sets the maximum number of entries to return from a search operation specifically which uses the simple paged results control. This overrides the nsslapd-sizelimit attribute for paged searches.

If this value is set to zero, then the nsslapd-sizelimit attribute is used for paged searches as well as non-paged searches.

This read-only attribute lists the DNs of the plug-in entries for the syntax and matching rule plug-ins loaded by the server.

Sets the bind DN used for an operation as the modifier of an entry, even if the operation itself was initiated by a server plug-in. The specific plug-in which performed the operation is listed in a separate operational attribute, internalModifiersname.

One change can trigger other, automatic changes in the directory tree. When a user is deleted, for example, that user is automatically removed from any groups it belonged to by the Referential Integrity Plug-in. The initial deletion of the user is performed by whatever user account is bound to the server, but the updates to the groups (by default) are shown as being performed by the plug-in, with no information about which user initiated that update. The nsslapd-plugin-binddn-tracking attribute allows the server to track which user originated an update operation, as well as the internal plug-in which actually performed it. For example:

dn: cn=my_group,ou=groups,dc=example,dc=com
modifiersname: uid=jsmith,ou=people,dc=example,dc=com
internalModifiersname: cn=referential integrity plugin,cn=plugins,cn=config

This attribute is disabled by default.

By default, even if access logging is set to record internal operations, plug-in internal operations are not logged in the access log file. Instead of enabling the logging in each plug-in’s configuration, you can control it globally with this parameter.

When enabled, plug-ins use this global setting and log access and audit events if enabled.

If nsslapd-plugin-logging is enabled and nsslapd-accesslog-level is set to record internal operations, unindexed searches and other internal operations are logged into the access log file.

In case nsslapd-plugin-logging is not set, unindexed searches from plug-ins are still logged in the Directory Server error log.

This attribute gives the TCP/IP port number used for standard LDAP communications. To run TLS over this port, use the Start TLS extended operation. This selected port must be unique on the host system; make sure no other application is attempting to use the same port number. Specifying a port number of less than 1024 means the Directory Server has to be started as root.

The server sets its uid to the nsslapd-localuser value after startup. When changing the port number for a configuration directory, the corresponding server instance entry in the configuration directory must be updated.

The server has to be restarted for the port number change to be taken into account.

This read-only attribute contains the list of the private naming contexts cn=config, cn=schema, and cn=monitor.

When the fine-grained password syntax is not set, new or updated passwords are not checked even though the global password syntax is configured. To inherit the global fine-grained password syntax, set this attribute to on.

Turns fine-grained (subtree- and user-level) password policy on and off.

If this attribute has a value of off, all entries (except for cn=Directory Manager) in the directory are subjected to the global password policy; the server ignores any defined subtree/user level password policy.

If this attribute has a value of on, the server checks for password policies at the subtree- and user-level and enforce those policies.

This attribute sets whether the whole server is in read-only mode, meaning that neither data in the databases nor configuration information can be modified. Any attempt to modify a database in read-only mode returns an error indicating that the server is unwilling to perform the operation.

This multi-valued attribute specifies the LDAP URLs to be returned by the suffix when the server receives a request for an entry not belonging to the local tree; that is, an entry whose suffix does not match the value specified on any of the suffix attributes. For example, assume the server contains only entries:

ou=People,dc=example,dc=com

but the request is for this entry:

ou=Groups,dc=example,dc=com

In this case, the referral would be passed back to the client in an attempt to allow the LDAP client to locate a server that contains the requested entry. Although only one referral is allowed per Directory Server instance, this referral can have multiple values.

For more information on managing referrals, see the "Configuring Directory Databases" chapter in the Red Hat Directory Server Administration Guide.

When set, this attribute sends back the referral for any request on any suffix.

This parameter requires that a user authenticate to the directory over a protected connection such as TLS, StartTLS, or SASL, rather than a regular connection.

This parameter lists what other core configuration attributes require that the server be restarted after a modification. This means that if any attribute listed in nsslapd-requiresrestart is changed, the new setting does not take effect until after the server is restarted. The list of attributes can be returned in an ldapsearch:

ldapsearch -D "cn=Directory Manager" -W -p 389 -h server.example.com -b "cn=config" -s sub -x "(objectclass=*)" | grep nsslapd-requiresrestart

This attribute is multi-valued.

This attribute specifies the number of file descriptors that Directory Server reserves for managing non-client connections, such as index management and managing replication. The number of file descriptors that the server reserves for this purpose subtracts from the total number of file descriptors available for servicing LDAP client connections (See Section 3.1.1.117, “nsslapd-maxdescriptors (Maximum File Descriptors)”).

Most installations of Directory Server should never need to change this attribute. However, consider increasing the value on this attribute if all of the following are true:

Increasing the value on this attribute may result in more LDAP clients being unable to access the directory. Therefore, the value on this attribute is increased, also increase the value on the nsslapd-maxdescriptors attribute. It may not be possible to increase the nsslapd-maxdescriptors value if the server is already using the maximum number of file descriptors that the operating system allows a process to use; see the operating system documentation for details. If this is the case, then reduce the load on the server by causing LDAP clients to search alternative directory replicas. See Section 3.1.1.61, “nsslapd-conntablesize” for information about file descriptor usage for incoming connections.

To assist in computing the number of file descriptors set for this attribute, use the following formula:

nsslapd-reservedescriptor = 20 + (NldbmBackends * 4) + NglobalIndex +
ReplicationDescriptor + ChainingBackendDescriptors + PTADescriptors + SSLDescriptors

The server has to be restarted for changes to this attribute to go into effect.

Returns the exact case of attribute type names as requested by the client. Although LDAPv3-compliant clients must ignore the case of attribute names, some client applications require attribute names to match exactly the case of the attribute as it is listed in the schema when the attribute is returned by the Directory Server as the result of a search or modify operation. However, most client applications ignore the case of attributes; therefore, by default, this attribute is disabled. Do not modify it unless there are legacy clients that can check the case of attribute names in results returned from the server.

The server has to be restarted for changes to this attribute to go into effect.

This attribute is deprecated and will be removed in a later version.

This attribute is used only for LDAPv2 clients that require attribute types to be returned with their RFC 1274 names. Set the value to on for those clients. The default is off.

This attribute sets the distinguished name (DN) of an entry that is not subject to access control restrictions, administrative limit restrictions for operations on the directory, or resource limits in general. There does not have to be an entry corresponding to this DN, and by default there is not an entry for this DN, thus values like cn=Directory Manager are acceptable.

For information on changing the root DN, see the "Creating Directory Entries" chapter in the Red Hat Directory Server Administration Guide.

This attribute sets the password associated with the Manager DN. When the root password is provided, it is encrypted according to the encryption method selected for the nsslapd-rootpwstoragescheme attribute. When viewed from the server console, this attribute shows the value *. When viewed from the dse.ldif file, this attribute shows the encryption method followed by the encrypted string of the password. The example shows the password as displayed in the dse.ldif file, not the actual password.

This attribute sets the method used to encrypt the Directory Server’s manager password stored in the nsslapd-rootpw attribute. For further details, such as recommended strong password storage schemes, see Section 4.1.43, “Password Storage Schemes”.

This setting does not require restarting the server to take effect.

This parameter sets the absolute path to the directory in which Directory Server stores run-time information, such as the PID file. The directory must be owned by the Directory Server user and group. Only this user and group must have read and write access in this directory.

The service must be restarted for changes to this attribute to take effect.

By default, only first matching SASL mapping is checked. If this mapping fails, the bind operation will fail even if there are other matching mappings that might have worked. SASL mapping fallback will keep checking all of the matching mappings.

You do not have to restart the server for this setting to take effect.

This attribute sets the maximum SASL buffer size.

Sets the absolute path to the directory containing the Cyrus-SASL SASL2 plug-ins. Setting this attribute allows the server to use custom or non-standard SASL plug-in libraries. This is usually set correctly during installation, and Red Hat strongly recommends not changing this attribute. If the attribute is not present or the value is empty, this means the Directory Server is using the system provided SASL plug-in libraries which are the correct version.

If this parameter is set, the server uses the specified path for loading SASL plug-ins. If this parameter is not set, the server uses the SASL_PATH environment variable. If neither nsslapd-saslpath or SASL_PATH are set, the server attempts to load SASL plug-ins from the default location, /usr/lib/sasl2.

Changes made to this attribute will not take effect until the server is restarted.

Ignores trailing spaces in object class names. By default, the attribute is turned off. If the directory contains entries with object class values that end in one or more spaces, turn this attribute on. It is preferable to remove the trailing spaces because the LDAP standards do not allow them.

For performance reasons, server restart is required for changes to take effect.

An error is returned by default when object classes that include trailing spaces are added to an entry. Additionally, during operations such as add, modify, and import (when object classes are expanded and missing superiors are added) trailing spaces are ignored, if appropriate. This means that even when nsslapd-schema-ignore-trailing-spaces is on, a value such as top is not added if top is already there. An error message is logged and returned to the client if an object class is not found and it contains trailing spaces.

This attribute sets whether the database schema is enforced when entries are added or modified. When this attribute has a value of on, Directory Server will not check the schema of existing entries until they are modified. The database schema defines the type of information allowed in the database. The default schema can be extended using the object classes and attribute types. For information on how to extend the schema using the Directory Server Console, see the "Extending the Directory Schema" chapter in the Red Hat Directory Server Administration Guide.

This is the absolute path to the directory containing the Directory Server instance-specific schema files. When the server starts up, it reads the schema files from this directory, and when the schema is modified through LDAP tools, the schema files in this directory are updated. This directory must be owned by the server user ID, and that user must have read and write permissions to the directory.

Changes made to this attribute will not take effect until the server is restarted.

Online schema modifications require a lock protection that are impacting the performance. If schema modifications are disabled, setting this parameter to off can increase the performance.

Determines whether modify operations that replace attribute values are allowed on the cn=schema entry.

If the attribute list passed to a search contains a space followed by other characters, the same string is returned to the client. For example:

# ldapsearch -b <basedn> "(filter)" "sn someothertext"
  dn: <matched dn>
  sn someothertext: <sn>

This behavior is disabled by default, but can be enabled using this configuration parameter.

This attribute allows multiple Directory Server instances to run on a multihomed machine (or makes it possible to limit listening to one interface of a multihomed machine). There can be multiple IP addresses associated with a single host name, and these IP addresses can be a mix of both IPv4 and IPv6. This parameter can be used to restrict the Directory Server instance to a single IP interface; this parameter also specifically sets what interface to use for TLS traffic rather than regular LDAP connections.

If a host name is given as the nsslapd-securelistenhost value, then the Directory Server responds to requests for every interface associated with the host name. If a single IP interface (either IPv4 or IPv6) is given as the nsslapd-securelistenhost value, Directory Server only responds to requests sent to that specific interface. Either an IPv4 or IPv6 address can be used.

The server has to be restarted for changes to this attribute to go into effect.

This attribute sets the TCP/IP port number used for TLS communications. This selected port must be unique on the host system; make sure no other application is attempting to use the same port number. Specifying a port number of less than 1024 requires that Directory Server be started as root. The server sets its uid to the nsslapd-localuser value after startup.

The server only listens to this port if it has been configured with a private key and a certificate, and nsslapd-security is set to on; otherwise, it does not listen on this port.

The server has to be restarted for the port number change to be taken into account.

This attribute sets whether the Directory Server is to accept TLS communications on its encrypted port. This attribute should be set to on for secure connections. To run with security on, the server must be configured with a private key and server certificate in addition to the other TLS configuration.

This attribute sets the maximum number of entries to return from a search operation. If this limit is reached, ns-slapd returns any entries it has located that match the search request, as well as an exceeded size limit error.

When no limit is set, ns-slapd returns every matching entry to the client regardless of the number found. To set a no limit value whereby the Directory Server waits indefinitely for the search to complete, specify a value of -1 for this attribute in the dse.ldif file.

This limit applies to everyone, regardless of their organization.

The corresponding user-level attribute is nsSizeLimit.

This parameter controls the SNMP index number of the Directory Server instance.

If you have multiple Directory Server instances on the same host listening all on port 389 but on different network interfaces, this parameter allows you to set different SNMP index numbers for each instance.

This attribute sets whether an TLS-enabled Directory Server should verify authenticity of a request by matching the host name against the value assigned to the common name (cn) attribute of the subject name (subjectDN field) in the certificate being presented. By default, the attribute is set to on. If it is on and if the host name does not match the cn attribute of the certificate, appropriate error and audit messages are logged.

For example, in a replicated environment, messages similar to the following are logged in the supplier server’s log files if it finds that the peer server’s host name does not match the name specified in its certificate:

[DATE] - SSL alert: ldap_sasl_bind("",LDAP_SASL_EXTERNAL) 81 (Netscape runtime error -12276 -
	 Unable to communicate securely with peer: requested domain name does not
	 match the server's certificate.)

[DATE] NSMMReplicationPlugin - agmt="cn=SSL Replication Agreement to host1" (host1.example.com:636):
 Replication bind with SSL client authentication failed:
 LDAP error 81 (Can't contact LDAP server)

Red Hat recommends turning this attribute on to protect Directory Server’s outbound TLS connections against a man in the middle (MITM) attack.

This attribute validates all modifications to entry attributes to make sure that the new or changed values conform to the required syntax for that attribute type. Any changes which do not conform to the proper syntax are rejected, when this attribute is enabled. All attribute values are validated against the syntax definitions in RFC 4514.

By default, this is turned on.

Syntax validation is only run against new or modified attributes; it does not validate the syntax of existing attribute values. Syntax validation is triggered for LDAP operations such as adds and modifies; it does not happen after operations like replication, since the validity of the attribute syntax should be checked on the originating supplier.

This validates all supported attribute types for Directory Server, with the exception of binary syntaxes (which cannot be verified) and non-standard syntaxes, which do not have a defined required format. The unvalidated syntaxes are as follows:

The nsslapd-syntaxcheck attribute sets whether to validate and reject attribute modifications. This can be used with the Section 3.1.1.167, “nsslapd-syntaxlogging” attribute to write warning messages about invalid attribute values to the error logs.

This attribute sets whether to log syntax validation failures to the errors log. By default, this is turned off.

If the Section 3.1.1.166, “nsslapd-syntaxcheck” attribute is enabled (the default) and the nsslapd-syntaxlogging attribute is also enabled, then any invalid attribute change is rejected and written to the errors log. If only nsslapd-syntaxlogging is enabled and nsslapd-syntaxcheck is disabled, then invalid changes are allowed to proceed, but a warning message is written to the error log.