Sunday 28 December 2014

nzbackup vs nz_backup

There are 2 commands to perform backup of Netezza.  One is nzbackup.  The other one is nz_backup.  For nzbackup, not much you can play with it except you need to specify the database to backup.  However, nz_backup provides more command options eg, specific table to backup, include where clause etc.  Here's the command options for nz_backup.


Usage:    nz_backup   -dir <dirname>  -format <ascii|binary|gzip>  [ optional args ]
               -or-
          nz_restore  -dir <dirname>  -format <ascii|binary|gzip>  [ optional args ]

Purpose:  To backup (or restore) one or more tables.

          An nz_backup must be run locally (on the NPS host being backed up).

          An nz_restore can be used to restore data into a remote NPS host.  Just
          include the "-host" switch, or set the NZ_HOST environment variable.

          Note:  When doing an "nz_restore -format binary -host <xxx>", the script
                 issues an "nzload -compress true".  This nzload feature only exists
                 in nzload as of NPS 4.6.  If you want to do this on an older version
                 of NPS (4.0 or 4.5) then:
                 o   Install a copy of the 4.6 client toolkit somewhere on your box
                     (it can be used against the older server releases)
                 o   Add it's bin directory to the start of your search PATH
                 o   Then invoke nz_restore

          These scripts can process a single table, multiple tables, or an entire database.

          The data format that is used can be either
               ascii  -- which is very portable.
               binary -- which is the database's compressed external table format.  This
                         is much faster, and results in significantly smaller backup sets.
               gzip   -- ascii, which is gzip'ed on the NPS host.

          The data can be written to (or read from) disk files or named pipes.
               If you use named pipes, another application is used to consume
               (or produce) the data.

          These scripts just concern themselves with the DATA itself.  When backing up
          a table, the DDL is not included.  When restoring a table, the script expects
          the table to already exist.  It will not create it.  It will not truncate it
          (so if the table currently has any data in it, that data will be left untouched
          by this script).

          To backup tables requires the following permissions:

               GRANT SELECT       ON  _VT_HOSTTXMGR         TO <user/group>;
               GRANT SELECT       ON  _VT_HOSTTX_INVISIBLE  TO <user/group>;
               GRANT SELECT       ON  _VT_DBOS_CONNECTION   TO <user/group>;
               --To obtain information about transactions

               GRANT LIST         ON  <DATABASE|dbname>     TO <user/group>;
               --The user must have access to the database that contains the tables

               GRANT SELECT       ON  <TABLE|tablename>     TO <user/group>;
               --The user must have access to the tables themselves, and their data

               GRANT CREATE EXTERNAL TABLE                  TO <user/group>;
               --The user must be able to create external tables, into which the
               --data will be unloaded

          To restore/reload a table requires an additional permission:

               GRANT INSERT       ON  <TABLE|tablename>     TO <user/group>;
               --The user must be able to insert (i.e., reload) data back into the tables

Options:  REQUIRED ARGUMENTS
          ==================

          -dir <dirname> [...]

               The full path to the directory in which the data files will be written
               to (or read from).  This directory must already exist and permit write
               access to it.  The directory name can be as meaningful as you wish to
               make it.

               If you are running this script as some linux user other than 'nz', please
               note that it will actually be one of the 'nz' processes that writes the
               data into this directory.  So linux user 'nz' must also have write access
               to it.  If you are using named pipes (rather than files) then this is the
               directory where the named pipes will be created.

               Examples:
                    -dir /backups/backup_set_17
                    -dir /snap_storage/bu/customer_db/2006_11_18
                    -dir /tmp

               If desired, you may split the backup files up across multiple directories/
               file systems.  Each thread can be associated with a separate "-dir <dirname>"
               by specifying them on the command line.  If you use this feature, then the
               number of directories specified must match the number of threads.

          -format <ascii|binary|gzip>

               Identifies the format to be used for the output files.

               ascii    Universal in nature, but typically results in larger files and
                        slower performance.

               binary   The database's compressed external table format.

               gzip     ascii, which is then compressed (using gzip).  By definition,
                        compressing and decompressing data uses up a lot of CPU cycles (i.e.,
                        it takes a long time).  When using the binary format (compressed/
                        external), the work is done in parallel across all of ths SPUs ... so
                        it is very quick.  But this option uses the NPS host to gzip/gunzip the
                        data.  You will (almost always) want to use multiple threads in order
                        to get more of the host's SMP processors involved in order to speed
                        things along.  The sweet spot seems to be about 8 threads, though you
                        can certainly use a larger/smaller number if you want to break the
                        backup files up into more/fewer pieces.

                                        best      --to--     least
                                       ======     ======     ======
                               Speed:  binary     ascii      gzip
                                Size:  gzip       binary     ascii
                        Universality:  ascii      gzip       binary


          OPTIONAL ARGUMENTS
          ==================

          -h or -? or --help       display this help

          -[rR]ev or -V            print the software revision of this program

          -host <name/IP>          host name or IP address  [NZ_HOST]
          -u    <user>             user name                [NZ_USER]
          -pw   <string>           password                 [NZ_PASSWORD]
          -db   <database>         database name            [NZ_DATABASE]

          -t <tablename> [...]

               # Table(s) within the database to be processed.  If none are specified,
               # then all tables in the database will be processed.
               #
               # If you have a file that contains a list of tablenames to be backed up,
               # (separated by spaces and/or newlines) it can be used via the following
               # syntax:
               #
               #      -t `cat /tmp/the_list_of_tables`

          -threads <nnn>

               # Breaking the backup into multiple threads (per table) can increase the
               # overall thruput, especially for large tables.  This will also have the
               # effect of creating smaller backup files, since each one will now be
               # broken up into multiple pieces.
               #
               # By default, only a single thread will be used.  You can specify a number
               # from 1..31.  Whatever value you specify for the backup must also be used
               # for the restore.  In general, the sweet spot seems to be about 6 threads.

          -script <scriptname>

               # Instead of backing up (or restoring) the data from disk files, you can use
               # named pipes -- allowing another application to consume (or produce) the data
               # on the fly.  To use named pipes, specify that application/script here.  The
               # script will be automatically invoked and passed (as arg 1) the fully rooted
               # pathname of the named pipe that is it supposed to use.
               #
               # For example scripts, see the file(s)
               #      nz_backup.script_example
               #      nz_restore.script_example

          -whereclause <clause>

               # Normally, nz_backup is used to backup the entire contents of a table --
               # all visible rows.  This option allows "you" to tack on a WHERE clause
               # to the data that gets selected ... allowing "you" to backup a subset
               # of the table.  All the power (and responsibility) is put into your
               # hands.  Do wrap the clause in double quotes so it will be passed into the
               # script correctly.  Examples:
               #
               #      -whereclause  "customer_key = 2"
               #      -whereclause  "customer_key in (1,3,5)"
               #      -whereclause  "region_name = 'AMERICA' or region_key = 0"
               #      -whereClause  "order_date between '1998-01-01' and '1998-12-31'"
               #
               # Because this clause gets applied to all tables being backed up, you would
               # probably only want to backup a single table at a time (when using this
               # clause) ... since the clause will typically contain column names that are
               # specific to that table.
               #
               # This clause only applies to backups (not restores).  Its use will be logged
               # in the output of this script (as well as in the pg.log file).

          -ignoreTxID

               # Tables are individually backed up, one at a time.  Since a backup may span
               # many hours, this script insures that the backup represents a consistent
               # point-in-time by using the transaction IDs attached to each row.
               #
               # This switch will override that feature ... and backs up each table with
               # whatever data it contains when the backup (of that particular table) is
               # kicked off.  This insures that your backup will include ALL of the data
               # in a table that has been committed (so you're not susceptible to long
               # running or stale transactions).
               #
               # This switch is primarily of use with NPS 4.0.  In later releases, this
               # script is able to do things differently.

          -dropdata

               # This is for testing purposes only.  As the name implies, the backup will
               # be written to /dev/null, resulting in no backup at all.  This is useful
               # for testing the performance of the NPS components that are involved
               # (SPUs/S-Blades ==> Host), while excluding the speed/overhead of your host
               # storage.

          -sizedata

               # This is for testing purposes only.  Like "-dropdata", but rather than
               # sending the backup data directly to /dev/null it will first be piped
               # thru "wc -c" in order to count the number of bytes in the backup
               # stream (e.g., to provide you with actual sizing information).  So it
               # has the performance characteristics of "-dropdata" ... but provides
               # you additional information.
               #
               # You can use "-format <ascii|binary>" and 1 or multiple "-threads <nn>"
               # when using this switch.
               #
               # Each table will include the following pieces of information
               #
               #      Info:  source table size       80,740,352
               #      Info:  backup file size        76,574,691
               #
               # And summary lines for the entire backup set will display the
               #
               #      TOTAL source table size :  48,267,526,144
               #      TOTAL backup file size  :  46,088,648,210
               #
               # The table size is whatever the table size is -- the amount of storage
               # space it using on disk (as reported by nz_db_size or nz_tables).  This
               # script doesn't know/care if the data is compressed on disk (e.g, CTA0/1/2).
               # Nor does the script know if there are any logically deleted rows in the
               # table (taking up space in the table, but which would not be part of a
               # backup data set).
               #
               # The backup size is the amount of storage that would be required if the
               # backup data set was actually written to disk.  This would represent either
               # the ascii or the binary (compressed external table format) version of the
               # data ... whatever you chose.
               #
               # To get sizing information for a full nzbackup, e.g.
               #      nzbackup   -db DBNAME  -dir /tmp
               # you would use a command line such as this
               #      nz_backup  -db DBNAME  -dir /tmp     -format binary  -sizedata

Outputs:  Status/log/timing information will be sent to standard out ... and will
          include information about any ERROR's that might be encountered.

          Exit status:  0 = success, non-0 = ERROR's were encountered

Examples: $ nz_backup  -format binary -dir /backupdir

          $ nz_restore -format ascii -dir /tmp -db my_db -t table1 -script /tmp/my_script

Comparison:               nzbackup/nzrestore     nz_backup/nz_restore
                          ==================     ====================
                           NPS CLI Utility          Add-on Script
Backup Type
 Full                                X                                 X
-differential                    X
-cumulative                    X

Granularity
Entire Database             X                                  X
  Individual Table(s)         nzrestore                   X

Formats Supported
Compressed Binary       X                                 X
  Ascii                             X
  Ascii gzip'ed                                                   X

Output Destination
  Veritas (NPS 4.0)               X
  Tivoli  (NPS 4.6)               X
  Disk                                   X                         X
  Named Pipes                                                 X

Multi-Stream Support         nzbackup (6.0)               X

DDL included as part              X              Use the nz_ddl* scripts
  of the backup set 

2 comments:

  1. Hi,

    can you please tell me the difference between nz_migrate and nz_backup/nz_restore ?

    ReplyDelete
  2. The comparison between "nzbackup" and "nzbackup" is an insightful read for database administrators. How Play Games This article's thorough analysis highlights the nuances and benefits of each tool.

    ReplyDelete