1<!--$Id: printlog.so,v 10.50 2007/10/26 15:39:18 bostic Exp $--> 2<!--Copyright (c) 1997,2008 Oracle. All rights reserved.--> 3<!--See the file LICENSE for redistribution information.--> 4<html> 5<head> 6<title>Berkeley DB Reference Guide: Reviewing Berkeley DB log files</title> 7<meta name="description" content="Berkeley DB: An embedded database programmatic toolkit."> 8<meta name="keywords" content="embedded,database,programmatic,toolkit,btree,hash,hashing,transaction,transactions,locking,logging,access method,access methods,Java,C,C++"> 9</head> 10<body bgcolor=white> 11<table width="100%"><tr valign=top> 12<td><b><dl><dt>Berkeley DB Reference Guide:<dd>Debugging Applications</dl></b></td> 13<td align=right><a href="../debug/runtime.html"><img src="../../images/prev.gif" alt="Prev"></a><a href="../toc.html"><img src="../../images/ref.gif" alt="Ref"></a><a href="../build_brew/intro.html"><img src="../../images/next.gif" alt="Next"></a> 14</td></tr></table> 15<p align=center><b>Reviewing Berkeley DB log files</b></p> 16<p>If you are running with transactions and logging, the <a href="../../utility/db_printlog.html">db_printlog</a> 17utility can be a useful debugging aid. The <a href="../../utility/db_printlog.html">db_printlog</a> utility 18will display the contents of your log files in a human readable (and 19machine-readable) format.</p> 20<p>The <a href="../../utility/db_printlog.html">db_printlog</a> utility will attempt to display any and all 21log files present in a designated db_home directory. For each log record, 22<a href="../../utility/db_printlog.html">db_printlog</a> will display a line of the form:</p> 23<blockquote><pre>[22][28]db_big: rec: 43 txnid 80000963 prevlsn [21][10483281]</pre></blockquote> 24<p>The opening numbers in square brackets are the <i>log sequence 25number</i> (<i>LSN</i>) of the log record being displayed. The first 26number indicates the log file in which the record appears, and the 27second number indicates the offset in that file of the record.</p> 28<p>The first character string identifies the particular log operation being 29reported. The log records corresponding to particular operations are 30described following. The rest of the line consists of name/value pairs.</p> 31<p>The rec field indicates the record type (this is used to dispatch records 32in the log to appropriate recovery functions).</p> 33<p>The txnid field identifies the transaction for which this record was 34written. A txnid of 0 means that the record was written outside the 35context of any transaction. You will see these most frequently for 36checkpoints.</p> 37<p>Finally, the prevlsn contains the LSN of the last record for this 38transaction. By following prevlsn fields, you can accumulate all the 39updates for a particular transaction. During normal abort processing, 40this field is used to quickly access all the records for a particular 41transaction.</p> 42<p>After the initial line identifying the record type, each field of the log 43record is displayed, one item per line. There are several fields that 44appear in many different records and a few fields that appear only in 45some records.</p> 46<p>The following table presents each currently written log record type with 47a brief description of the operation it describes. Any of these 48record types may have the string "_debug" appended if they 49were written because <a href="../../api_c/db_set_flags.html#DB_TXN_NOT_DURABLE">DB_TXN_NOT_DURABLE</a> was specified and the 50system was configured with <a href="../../ref/build_unix/conf.html#--enable-diagnostic">--enable-diagnostic</a>.</p> 51<!--START LOG RECORD TYPES--> 52<table border=1> 53<tr><th>Log Record Type</th><th>Description</th></tr> 54<tr><td>bam_adj</td><td>Used when we insert/remove an index into/from the page header of a Btree page.</td></tr> 55<tr><td>bam_cadjust</td><td>Keeps track of record counts in a Btree or Recno database.</td></tr> 56<tr><td>bam_cdel</td><td>Used to mark a record on a page as deleted.</td></tr> 57<tr><td>bam_curadj</td><td>Used to adjust a cursor location when a nearby record changes in a Btree database.</td></tr> 58<tr><td>bam_merge</td><td>Used to merge two Btree database pages during compaction.</td></tr> 59<tr><td>bam_pgno</td><td>Used to replace a page number in a Btree record.</td></tr> 60<tr><td>bam_rcuradj</td><td>Used to adjust a cursor location when a nearby record changes in a Recno database.</td></tr> 61<tr><td>bam_relink</td><td>Fix leaf page prev/next chain when a page is removed.</td></tr> 62<tr><td>bam_repl</td><td>Describes a replace operation on a record.</td></tr> 63<tr><td>bam_root</td><td>Describes an assignment of a root page.</td></tr> 64<tr><td>bam_rsplit</td><td>Describes a reverse page split.</td></tr> 65<tr><td>bam_split</td><td>Describes a page split.</td></tr> 66<tr><td>crdel_inmem_create</td><td>Record the creation of an in-memory named database.</td></tr> 67<tr><td>crdel_inmem_remove</td><td>Record the removal of an in-memory named database.</td></tr> 68<tr><td>crdel_inmem_rename</td><td>Record the rename of an in-memory named database.</td></tr> 69<tr><td>crdel_metasub</td><td>Describes the creation of a metadata page for a subdatabase.</td></tr> 70<tr><td>db_addrem</td><td>Add or remove an item from a page of duplicates.</td></tr> 71<tr><td>db_big</td><td>Add an item to an overflow page (<i>overflow pages</i> contain items too large to place on the main page)</td></tr> 72<tr><td>db_cksum</td><td>Unable to checksum a page.</td></tr> 73<tr><td>db_debug</td><td>Log debugging message.</td></tr> 74<tr><td>db_noop</td><td>This marks an operation that did nothing but update the LSN on a page.</td></tr> 75<tr><td>db_ovref</td><td>Increment or decrement the reference count for a big item.</td></tr> 76<tr><td>db_pg_alloc</td><td>Indicates we allocated a page to a database.</td></tr> 77<tr><td>db_pg_free</td><td>Indicates we freed a page (freed pages are added to a freelist and reused).</td></tr> 78<tr><td>db_pg_freedata</td><td>Indicates we freed a page that still contained data entries (freed pages are added to a freelist and reused.)</td></tr> 79<tr><td>db_pg_init</td><td>Indicates we reinitialized a page during a truncate.</td></tr> 80<tr><td>db_pg_sort</td><td>Sort the free page list and free pages at the end of the file.</td></tr> 81<tr><td>dbreg_register</td><td>Records an open of a file (mapping the filename to a log-id that is used in subsequent log operations).</td></tr> 82<tr><td>fop_create</td><td>Create a file in the file system.</td></tr> 83<tr><td>fop_file_remove</td><td>Remove a name in the file system.</td></tr> 84<tr><td>fop_remove</td><td>Remove a file in the file system.</td></tr> 85<tr><td>fop_rename</td><td>Rename a file in the file system.</td></tr> 86<tr><td>fop_write</td><td>Write bytes to an object in the file system.</td></tr> 87<tr><td>ham_chgpg</td><td>Used to adjust a cursor location when a Hash page is removed, and its elements are moved to a different Hash page.</td></tr> 88<tr><td>ham_copypage</td><td>Used when we empty a bucket page, but there are overflow pages for the bucket; one needs to be copied back into the actual bucket.</td></tr> 89<tr><td>ham_curadj</td><td>Used to adjust a cursor location when a nearby record changes in a Hash database.</td></tr> 90<tr><td>ham_groupalloc</td><td>Allocate some number of contiguous pages to the Hash database.</td></tr> 91<tr><td>ham_insdel</td><td>Insert/delete an item on a Hash page.</td></tr> 92<tr><td>ham_metagroup</td><td>Update the metadata page to reflect the allocation of a sequence of contiguous pages.</td></tr> 93<tr><td>ham_newpage</td><td>Adds or removes overflow pages from a Hash bucket.</td></tr> 94<tr><td>ham_replace</td><td>Handle updates to records that are on the main page.</td></tr> 95<tr><td>ham_splitdata</td><td>Record the page data for a split.</td></tr> 96<tr><td>qam_add</td><td>Describes the actual addition of a new record to a Queue.</td></tr> 97<tr><td>qam_del</td><td>Delete a record in a Queue.</td></tr> 98<tr><td>qam_delext</td><td>Delete a record in a Queue with extents.</td></tr> 99<tr><td>qam_incfirst</td><td>Increments the record number that refers to the first record in the database.</td></tr> 100<tr><td>qam_mvptr</td><td>Indicates we changed the reference to either or both of the first and current records in the file.</td></tr> 101<tr><td>txn_child</td><td>Commit a child transaction.</td></tr> 102<tr><td>txn_ckp</td><td>Transaction checkpoint.</td></tr> 103<tr><td>txn_recycle</td><td>Transaction IDs wrapped.</td></tr> 104<tr><td>txn_regop</td><td>Logs a regular (non-child) transaction commit.</td></tr> 105<tr><td>txn_xa_regop</td><td>Logs a prepare message.</td></tr> 106</table> 107<!--END LOG RECORD TYPES--> 108<b>Augmenting the Log for Debugging</b> 109<p>When debugging applications, it is sometimes useful to log not only the 110actual operations that modify pages, but also the underlying Berkeley DB 111functions being executed. This form of logging can add significant bulk 112to your log, but can permit debugging application errors that are almost 113impossible to find any other way. To turn on these log messages, specify 114the --enable-debug_rop and --enable-debug_wop configuration options when 115configuring Berkeley DB. See <a href="../../ref/build_unix/conf.html">Configuring 116Berkeley DB</a> for more information.</p> 117<b>Extracting Committed Transactions and Transaction Status</b> 118<p>Sometimes, it is helpful to use the human-readable log output to 119determine which transactions committed and aborted. The awk script, 120commit.awk, (found in the db_printlog directory of the Berkeley DB 121distribution) allows you to do just that. The following command, 122where log_output is the output of db_printlog, will display a list of 123the transaction IDs of all committed transactions found in the log:</p> 124<blockquote><pre>awk -f commit.awk log_output</pre></blockquote> 125<p>If you need a complete list of both committed and aborted transactions, 126then the script status.awk will produce it. The syntax is as follows:</p> 127<blockquote><pre>awk -f status.awk log_output</pre></blockquote> 128<b>Extracting Transaction Histories</b> 129<p>Another useful debugging aid is to print out the complete history of a 130transaction. The awk script txn.awk allows you to do that. The 131following command line, where log_output is the output of 132<a href="../../utility/db_printlog.html">db_printlog</a> and txnlist is a comma-separated list of transaction 133IDs, will display all log records associated with the designated 134transaction ids:</p> 135<blockquote><pre>awk -f txn.awk TXN=txnlist log_output</pre></blockquote> 136<b>Extracting File Histories</b> 137<p>The awk script fileid.awk allows you to extract all log records that 138refer to a designated file. The syntax for the fileid.awk script is 139the following, where log_output is the output of db_printlog and fids 140is a comma-separated list of fileids:</p> 141<blockquote><pre>awk -f fileid.awk PGNO=fids log_output</pre></blockquote> 142<b>Extracting Page Histories</b> 143<p>The awk script pgno.awk allows you to extract all log records that refer 144to designated page numbers. However, because this script will extract 145records with the designated page numbers for all files, it is most 146useful in conjunction with the fileid script. The syntax for the 147pgno.awk script is the following, where log_output is the output of 148db_printlog and pgnolist is a comma-separated list of page numbers:</p> 149<blockquote><pre>awk -f pgno.awk PGNO=pgnolist log_output</pre></blockquote> 150<b>Other log processing tools</b> 151<p>The awk script count.awk prints out the number of log records 152encountered that belonged to some transaction (that is, the number of 153log records excluding those for checkpoints and 154non-transaction-protected operations).</p> 155<p>The script range.awk will extract a subset of a log. This is useful 156when the output of <a href="../../utility/db_printlog.html">db_printlog</a> is too large to be reasonably 157manipulated with an editor or other tool. The syntax for range.awk is 158the following, where <b>sf</b> and <b>so</b> represent the LSN 159of the beginning of the sublog you want to extract, and <b>ef</b> and 160<b>eo</b> represent the LSN of the end of the sublog you want to 161extract:</p> 162<blockquote><pre>awk -f range.awk START_FILE=sf START_OFFSET=so END_FILE=ef END_OFFSET=eo log_output</pre></blockquote> 163<table width="100%"><tr><td><br></td><td align=right><a href="../debug/runtime.html"><img src="../../images/prev.gif" alt="Prev"></a><a href="../toc.html"><img src="../../images/ref.gif" alt="Ref"></a><a href="../build_brew/intro.html"><img src="../../images/next.gif" alt="Next"></a> 164</td></tr></table> 165<p><font size=1>Copyright (c) 1996,2008 Oracle. All rights reserved.</font> 166</body> 167</html> 168