Document Type | Technical Information
Category | Administration
Applicable Product Versions | 6FS07PS, 7FS02PS
Document Number | TADTI008
Overview
tblogdump is a utility that can generate a dump when Logfile Dump cannot be used due to the database being unable to start.
tblogdump Limitations
- tblogdump cannot read encrypted redo logs.
- tblogdump cannot be used when Transparent Tablespace Encryption (TTE) is enabled.
- Extracting large log dumps using tblogdump on NFS may take a long time.
Method
How to Replace tblogdump
tblogdump can be downloaded from Technet and replaced in the $TB_HOME/client/bin directory for use.
- Access Tibero Technet - Technet > Download > Database > Tibero > Tibero Trouble Shooting Tools > Download
- Select platform for Tibero Trouble Shooting Tools > Linux > Download tblogdump
- Access Tibero server > navigate to the directory below, upload the downloaded file and replace it
-. Backup existing file $ cp $TB_HOME/client/bin/tblogdump $TB_HOME/client/bin/tblogdump.bak -. Overwrite with new file $ mv tblogdump $TB_HOME/client/bin/ $ chmod +x $TB_HOME/client/bin/tblogdump
How to Use tblogdump
tblogdump usage examples are as follows.
$ tblogdump [options] <redo log file path>
$ tblogdump Usage: tblogdump [options] <fpath> <fpath> fullpath of redo log files to dump Options: -a | --dba specifies dba (data block address) -n log block number (format: 5 or 3-7, 0: header blk) -s log block size (default: 512) --seq log sequence number (format: 5 or 3-7) must be used with --thread option --opcode specifies opcode (format: opcode1.opcode2) --sgmtid specifies segment id --xid specifies transaction id (format: xxxx.xx.xxxx) must be used with --seq option --thread specifies redo thread id (only valid with --seq option) --unique specifies when the index is a unique index --nonunique specifies when the index is a nonunique index --kd_len specifies kd_len of irp --keycolcnt specifies keycolcnt of irp --rowid specifies rowid (format: dba.rowno or sgmtid.dba.rowno) --over-nab targets all the blocks over the nab but only within the same log sequence. --tsn specifies tsn (format: xxxx.xxxxxxxx(hex) or decimal value, single value or range(a-b)) Example: tblogdump $TB_HOME/database/$TB_SID/log001.log tblogdump --seq 3-5 --thread 0 $TB_HOME/database/$TB_SID/
tblogdump Options
How to Use tblogdump by Case
Dump Entire Log File
$ tblogdump <file>
Dump Log File Header Block
$ tblogdump -n 0 <file>
Dump Logs for a Specific Block Only
$ tblogdump -n <block_no> <file>
Dump Logs for a Specific DBA (data block address)
$ tblogdump -a <dba> <file>
Dump Logs Using a Range of Log File Sequences as Input
$ tblogdump --seq <seq/seq start-end> <dir_path>
Dump Logs Using a Specific Opcode from the Log File as Input
$ tblogdump --opcode <major opcode number.minor opcode number> <file>
Dump Logs for a Specific sgmtid in the Log File
$ tblogdump --sgmtid <segment id> <File>
Dump Logs for a Specific Transaction ID in the Log File
$ tblogdump --seq <sequence> --xid <xid> --thread <thread_id>
Dump Logs for a Specific Thread in the Log File
$ tblogdump --seq <seq> --thread <thread> <dir_path>
Dump Logs for a Specific tsn in the Log File
$ tblogdump --tsn <tsn> <file>