vserver services ndmp modify

Modify NDMP Properties

Availability: This command is available to cluster and Vserver administrators at the admin privilege level.

Description

This command is used to change NDMP options on Vservers.

One or more of the options specified in the parameters section can be modified for a specific Vserver, by this command. A short description of each of the options is provided in the parameters section.

Parameters

-vserver <Vserver Name> - Vserver
This parameter specifies the name of the Vserver.
[-ignore-ctime-enabled {true|false}] - Ignore Ctime

This option, when true, allows user to exclude files with ctime changed from storage system' incremental dumps since other processes like virus scanning often alter the ctime of files. When this option is false, backup on the Vserver will include all files with a change or modified time later then the last dump in the previous level dump. The default value is false. This option is persistent across reboots.

Most WIN32 APIs are often unaware of the "last changed time", ctime, they often incorrectly set a later time for files, causing these files to be included in the Vserver's incremental dumps, making the incremental dump very large. This is partially defying the purpose of having incremental dumps, since one uses incremental dumps to speed up the backup by only dumping files that were truly changed since the last backup.

The -option-value for this parameter should be true/false.

[-offset-map-enable {true|false}] - Enable Offset Map

This option is used to enable or disable generation of the inode offset map during NDMP based dump backups. The offset map is required to perform Enhanced Direct Access Restore (DAR) on the backup data. Enhanced DAR provides support for directory DAR and DAR of files with NT streams. The default value for this option is true. This option is persistent across reboots.

The -option-value for this parameter should be true/false.

[-tcpnodelay {true|false}] - Enable TCP Nodelay

Enables/Disables the TCPNODELAY configuration parameter for the socket between the Vserver and the DMA. When set to true, the Nagle algorithm is disabled and small packets are sent immediately rather than held and bundled with other small packets. This optimizes the system for response time rather than throughput.

This option becomes active when the next NDMP session starts. Existing sessions are unaffected. The default value for this option is false. This option is persistent across reboots.

The -option-value for this parameter should be true/false.

[-tcpwinsize <integer>] - TCP Window Size

This option can be used to change the TCP buffer size of the NDMP data connection.The minimum and maximum values are 8192(8K) and 262,144(256K), respectively. The default value for this option is 32768(32K).

This option is persistent across reboots.

The -option-value for this parameter should be a number between 8192(8K) and 262,144(256K).

[-data-port-range <text>] - Data Port Range

This option allows administrators to specify a port range on which the NDMP server can listen for data connections.

The format of this option is start_port - end_port start_port, end_port can have values between [1024-65535]; start_port must be lesser than or equal to end_port. If a valid range is specified, NDMP uses a port within that range to listen for data connections. A listen request fails if no ports in the specified range are free.

This option is modifiable only from the admin Vserver context and the said option is applicable for all the data Vservers and the admin Vserver. For example, if the value of the above option is set with 2000-3000, the same value will be applicable throughout the cluster. The value all implies that any available port can be used to listen for data connections. The default value for this option is all. This option is persistent across reboots.

The -option-value for this option should be in the format {<start_port>-<end port> | all }- where start_port, end_port can have values between [1024-65535]; start_port must be lesser than or equal to end_port.

[-backup-log-enable {true|false}] - Enable Backup Log

Backup logging captures important events during dump/restore and records them in /mroot/etc/log/backup on the root volume. The option allows users to enable or disable this feature. The default value for this option is true. This option is persistent across reboots.

The -option-value for this parameter should be true/false.

[-per-qtree-exclude-enable {true|false}] - Enable per Qtree Exclusion

If this option is true, users can specify exclude list on a per qtree basis to be excluded from backup. This exclude list will override any values already present due to 'EXCLUDE' environment variable . The user can specify the exclusion list through a .exclude_list file which resides at the root of the qtree. The exclusion list can be a list of files or files that match a specified pattern. The default value for this option is false. This option is persistent across reboots.

The -option-value for this parameter should be true/false.

[-authtype <NDMP Authentication types>, ...] - Authentication Type

Allows the administrator to choose the authentication method. NDMP supports three authentication types: challenge, plaintext and plaintext_sso. The plaintext_sso authentication type is mutually exclusive with the other authentication types. By setting the authentication type as plaintext_sso, the actual password for the user can be used to authenticate instead of having to generate an NDMP specific password. The default of this option is challenge. This option is persistent across reboots.

The -option-value for this parameter can be {challenge | plaintext | plaintext_sso | challenge, plaintext | plaintext, challenge}.

[-debug-enable {true|false}] - Enable Debug (privilege: advanced)

This option enables debug logging for NDMP. Debug messages will be logged to the ndmpd log file /mroot/etc/log/mlog/ndmpd.log . The default value for this option is false .This option is persistent across reboots.

The -option-value for this parameter should be true/false.

[-debug-filter <text>] - Debug Filter (privilege: advanced)

This option controls the NDMP modules for which debug logging is to be enabled. option-value can take five values for this option : all, none, normal, backend or "filter-expression".

all enables debug logging for all modules.

none disables debug logging for all modules. It is equivalent to modify -vserver vserver_name -debug-enable false.

normal is a shortcut option that enables debug logging for all modules except verbose and io_loop. The equivalent filter string is all-verbose-io_loop.

backend is a short cut option that enables debug logging for all modules except verbose, io_loop, ndmps and ndmpd. The equivalent filter string is all-verbose-io_loop-ndmps-ndmpp.

(filter-expression) is a combination of one or more modules for which debug logs needs to be enabled. Multiple module names can be combined using following operators :

  • - to remove the given module from the list of specified modules in the filter string. For example the filter all-ndmpp will enable debug logging for all modules but not ndmpp.
  • ^ to add the given module or modules to the list of modules specified in the filter string. For example the filter ndmpp^mover^data will enable debug logging for ndmpp, mover and data.
The possible module names and a brief description is given below:-
        +--------------+-------------------------------------+
        | Modules      | Description                         |
        +--------------+-------------------------------------+
        | verbose      | verbose message                     |
        | io           | I/O process loop                    |
        | io_loop      | I/O process loop verbose messages   |
        | ndmps        | NDMP service                        |
        | ndmpp        | NDMP Protocol                       |
        | rpc          | General RPC service                 |
        | fdc_rpc      | RPC to FC driver service            |
        | auth         | Authentication                      |
        | mover        | NDMP MOVER (tape I/O)               |
        | data         | NDMP DATA (backup/restore)          |
        | scsi         | NDMP SCSI (robot/tape ops)          |
        | bkup_rpc     | RPC to Backup service client        |
        | bkup_rpc_s   | RPC to Backup service server        |
        | conf         | Debug configure/reconfigure         |
        | dblade       | Dblade specific messages            |
        | timer        | NDMP server timeout messages        |
        | vldb         | VLDB service                        |
        | smf          | SMF Gateway messages                |
        | common       | NDMP common state                   |
        | ext          | NDMP extensions messages            |
        | ndmprpc      | NDMP Mhost RPC server               |
        +--------------+-------------------------------------+
			

The default value for this option is none. This option is persistent across reboots.

The -option-value for this parameter can be {all | none | normal | backend |'filter-expression'}.

[-dump-logical-find <text>] - Enable Logical Find for Dump (privilege: advanced)

This option specifies whether to follow inode-file walk or tree walk for phase I of the dump. Choosing inode-file walk or tree walk affects the performance of the dump. This option can take following values:

If default is specified, then level 0 and incremental volume as well as qtree dumps will use inode walk. All the subtree dumps will use tree walk.

If always is specified, all dumps will follow treewalk.

A comma-separated list of values in any combination from the following list:

  • vol_baseline: Level 0 full volume backup will follow treewalk.
  • vol_incr: Incremental full volume backup will follow treewalk.
  • qtree_baseline: Level 0 qtree backup will follow treewalk.
  • qtree_incr: Incremental qtree backup will follow treewalk.

The default value for this option is default. This option is persistent across reboots.

The -option-value for this parameter could be {default | always | 'vol_baseline' | 'vol_baseline,qtree_baseline' | ...}.

[-abort-on-disk-error {true|false}] - Enable Abort on Disk Error (privilege: advanced)

If this option is true, dump will abort the backup operation on detection of irrecoverable data blocks in user files. If this option is false, dump will proceed with backup operation - even if irrecoverable data blocks in user files are detected. On detection of irrecoverable data blocks, dump will send a log message to DMA and also log an entry in /mroot/etc/log/backup file. The default value for this option is false. This option is persistent across reboots.

The -option-value for this parameter should be true/false.

[-fh-dir-retry-interval <integer>] - FH Throttle Value for Dir (privilege: advanced)

NDMP protocol sends back file history information for all directories in phase 3 of dump to DMA. In the presence of slow DMA or high latency networks, the amount of file history being generated exceeds the amount being consumed by the DMA. To handle a slow reader, a flow control mechanism is now introduced where file history generation is throttled when a DMA is slow in consuming them. The value for this option indicates how frequently should the file history be resent if it was throttled. The default value is 250 milliseconds. This option is persistent across reboots.

The -option-value for this parameter should be a number.

[-fh-node-retry-interval <integer>] - FH Throttle Value for Node (privilege: advanced)

NDMP protocol sends back file history information for all files in phase 4 of dump to DMA. In the presence of slow DMA or high latency networks, the amount of file history being generated exceeds the amount being consumed by the DMA. To handle slow reader conditions, a flow control mechanism is now introduced where file history generation is throttled when a DMA is slow in consuming them. The value for this option indicates how frequently should the file history be resent if it was throttled. The default value is 250 milliseconds. This option is persistent across reboots.

The -option-value for this parameter should be a number.

[-restore-vm-cache-size <integer>] - Restore VM File Cache Size (privilege: advanced)

This option mandates the number of WAFL buffers pinned in memory by various meta-files used by logical restore. The minimum and maximum values are 4 and 1024, respectively. The default value for this option is 64. This option is persistent across reboots.

Depending on the value of this option, various meta-files are assigned a number of WAFL buffers that need to be pinned in memory.

        +-----------------+---------------------------------+
        | Meta-filename   |     Number of WAFL buffers to   |
        |                 |      be pinned in memory        |
        +-----------------+---------------------------------+
        | dumpmap         | ndmpd.restore.vm_cache_size     |
        | filemap         | ndmpd.restore.vm_cache_size     |
        | aclfile_map     | ndmpd.restore.vm_cache_size     |
        | inomap          | ndmpd.restore.vm_cache_size / 2 |
        | basemap         | ndmpd.restore.vm_cache_size / 2 |
        | flipmap         | ndmpd.restore.vm_cache_size / 2 |
        | revmap          | ndmpd.restore.vm_cache_size / 2 |
        | clrimap         | ndmpd.restore.vm_cache_size / 4 |
        | mfp_for_inotab  | ndmpd.restore.vm_cache_size / 4 |
        | map             | ndmpd.restore.vm_cache_size / 4 |
        | offsetfile_map  | ndmpd.restore.vm_cache_size / 4 |
        +-----------------+---------------------------------+
			

The -option-value for this parameter should be a number between 4 and 1024.

[-enable {true|false}] - Enable NDMP on Vserver

When the option is set to true, the NDMP daemon handles requests, and when set to false, the NDMP daemon does not handle requests. Enabling and disabling the option is equivalent to executing the following commands: vserver services ndmp on and vserver services ndmp off respectively. This option is persistent across reboots. The default value of this option is false.

The -option-value for this parameter is either true or false.

[-preferred-interface-role {cluster|data|node-mgmt|intercluster|cluster-mgmt}, ...] - Preferred Interface Role

This option allows the user to specify the preferred Logical Interface (LIF) role while establishing an NDMP data connection channel. The NDMP data server or the NDMP mover establishes a data channel from the node that owns the volume or the tape device respectively. This option is used on the node that owns the volume or the tape device. The order of IP addresses that are used to establish the data connection depends on the order of LIF roles specified in this option.

The default value for this option for the admin Vserver is intercluster, cluster-mgmt, node-mgmt

The default value for this option for a data Vserver is intercluster, data.

[-secondary-debug-filter <text>] - Secondary Debug Filter (privilege: advanced)

This option allows control on NDMP debug logging. This option takes a comma separated tag=value pairs. The supported tag is IPADDR which can be used to specify Vserver IP addresses for which NDMP debugging is required. If this option is set and the option debug-enable is set to true, then the debug-filter option is applicable to sessions whose control connection IP addresses match the IP addresses that are listed in the option. If this option is not set, the debug filter is applicable to all Vserver sessions. By default, this option does not have a value set.

[-is-secure-control-connection-enabled {true|false}] - Is Secure Control Connection Enabled

This option enables NDMP service to accept control connections over secure sockets on TCP port 30000. This option is persistent across reboots. The default value of this option is false.

Examples

The following example show how to enable NDMP on a Vserver and set authorization type to plaintext :
cluster1::> vserver services ndmp modify -vserver vs1 -enable true -authtype plaintext
cluster1::>