21.5.23 ndb_redo_log_reader — Check and Print Content of Cluster Redo Log

Reads a redo log file, checking it for errors, printing its contents in a human-readable format, or both. ndb_redo_log_reader is intended for use primarily by NDB Cluster developers and Support personnel in debugging and diagnosing problems.

This utility remains under development, and its syntax and behavior are subject to change in future NDB Cluster releases.

The C++ source files for ndb_redo_log_reader can be found in the directory /storage/ndb/src/kernel/blocks/dblqh/redoLogReader.

Options that can be used with ndb_redo_log_reader are shown in the following table. Additional descriptions follow the table.

Table 21.37 Command-line options used with the program ndb_redo_log_reader

Format	Description	Added, Deprecated, or Removed
-dump	Print dump info	(Supported in all NDB releases based on MySQL 5.7)
-filedescriptors	Print file descriptors only	(Supported in all NDB releases based on MySQL 5.7)
--help	Print usage information (has no short form)	(Supported in all NDB releases based on MySQL 5.7)
-lap	Provide lap info, with max GCI started and completed	(Supported in all NDB releases based on MySQL 5.7)
-mbyte #	Starting megabyte	(Supported in all NDB releases based on MySQL 5.7)
-mbyteheaders	Show only first page header of each megabyte in file	(Supported in all NDB releases based on MySQL 5.7)
-nocheck	Do not check records for errors	(Supported in all NDB releases based on MySQL 5.7)
-noprint	Do not print records	(Supported in all NDB releases based on MySQL 5.7)
-page #	Start with this page	(Supported in all NDB releases based on MySQL 5.7)
-pageheaders	Show page headers only	(Supported in all NDB releases based on MySQL 5.7)
-pageindex #	Start with this page index	(Supported in all NDB releases based on MySQL 5.7)
-twiddle	Bit-shifted dump	(Supported in all NDB releases based on MySQL 5.7)

Usage

ndb_redo_log_reader file_name [options]

file_name is the name of a cluster redo log file. redo log files are located in the numbered directories under the data node's data directory (DataDir); the path under this directory to the redo log files matches the pattern ndb_nodeid_fs/D#/DBLQH/S#.FragLog. nodeid is the data node's node ID. The two instances of # each represent a number (not necessarily the same number); the number following D is in the range 8-39 inclusive; the range of the number following S varies according to the value of the NoOfFragmentLogFiles configuration parameter, whose default value is 16; thus, the default range of the number in the file name is 0-15 inclusive. For more information, see NDB Cluster Data Node File System Directory.

The name of the file to be read may be followed by one or more of the options listed here:

• -dump

  Command-Line Format	-dump

  Print dump info.

• Command-Line Format	-filedescriptors

  -filedescriptors: Print file descriptors only.

• Command-Line Format	--help

  --help: Print usage information.

• -lap

  Command-Line Format	-lap

  Provide lap info, with max GCI started and completed.

• Command-Line Format	-mbyte #
  Type	Numeric
  Default Value	0
  Minimum Value	0
  Maximum Value	15

  -mbyte #: Starting megabyte.

  # is an integer in the range 0 to 15, inclusive.

• Command-Line Format	-mbyteheaders

  -mbyteheaders: Show only the first page header of every megabyte in the file.

• Command-Line Format	-noprint

  -noprint: Do not print the contents of the log file.

• Command-Line Format	-nocheck

  -nocheck: Do not check the log file for errors.

• Command-Line Format	-page #
  Type	Integer
  Default Value	0
  Minimum Value	0
  Maximum Value	31

  -page #: Start at this page.

  # is an integer in the range 0 to 31, inclusive.

• Command-Line Format	-pageheaders

  -pageheaders: Show page headers only.

• Command-Line Format	-pageindex #
  Type	Integer
  Default Value	12
  Minimum Value	12
  Maximum Value	8191

  -pageindex #: Start at this page index.

  # is an integer between 12 and 8191, inclusive.

• -twiddle

  Command-Line Format	-twiddle

  Bit-shifted dump.

Like ndb_print_backup_file and ndb_print_schema_file (and unlike most of the NDB utilities that are intended to be run on a management server host or to connect to a management server) ndb_redo_log_reader must be run on a cluster data node, since it accesses the data node file system directly. Because it does not make use of the management server, this utility can be used when the management server is not running, and even when the cluster has been completely shut down.