1<?xml version="1.0" encoding="UTF-8" standalone="no"?> 2<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> 3<html xmlns="http://www.w3.org/1999/xhtml"> 4 <head> 5 <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /> 6 <title>Chapter 5. Managing DB Files</title> 7 <link rel="stylesheet" href="gettingStarted.css" type="text/css" /> 8 <meta name="generator" content="DocBook XSL Stylesheets V1.62.4" /> 9 <link rel="home" href="index.html" title="Getting Started with Berkeley DB Transaction Processing" /> 10 <link rel="up" href="index.html" title="Getting Started with Berkeley DB Transaction Processing" /> 11 <link rel="previous" href="reversesplit.html" title="Reverse BTree Splits" /> 12 <link rel="next" href="backuprestore.html" title="Backup Procedures" /> 13 </head> 14 <body> 15 <div class="navheader"> 16 <table width="100%" summary="Navigation header"> 17 <tr> 18 <th colspan="3" align="center">Chapter 5. Managing DB Files</th> 19 </tr> 20 <tr> 21 <td width="20%" align="left"><a accesskey="p" href="reversesplit.html">Prev</a> </td> 22 <th width="60%" align="center"> </th> 23 <td width="20%" align="right"> <a accesskey="n" href="backuprestore.html">Next</a></td> 24 </tr> 25 </table> 26 <hr /> 27 </div> 28 <div class="chapter" lang="en" xml:lang="en"> 29 <div class="titlepage"> 30 <div> 31 <div> 32 <h2 class="title"><a id="filemanagement"></a>Chapter 5. Managing DB Files</h2> 33 </div> 34 </div> 35 <div></div> 36 </div> 37 <div class="toc"> 38 <p> 39 <b>Table of Contents</b> 40 </p> 41 <dl> 42 <dt> 43 <span class="sect1"> 44 <a href="filemanagement.html#checkpoints">Checkpoints</a> 45 </span> 46 </dt> 47 <dt> 48 <span class="sect1"> 49 <a href="backuprestore.html">Backup Procedures</a> 50 </span> 51 </dt> 52 <dd> 53 <dl> 54 <dt> 55 <span class="sect2"> 56 <a href="backuprestore.html#copyutilities">About Unix Copy Utilities</a> 57 </span> 58 </dt> 59 <dt> 60 <span class="sect2"> 61 <a href="backuprestore.html#standardbackup">Offline Backups</a> 62 </span> 63 </dt> 64 <dt> 65 <span class="sect2"> 66 <a href="backuprestore.html#hotbackup">Hot Backup</a> 67 </span> 68 </dt> 69 <dt> 70 <span class="sect2"> 71 <a href="backuprestore.html#incrementalbackups">Incremental Backups</a> 72 </span> 73 </dt> 74 </dl> 75 </dd> 76 <dt> 77 <span class="sect1"> 78 <a href="recovery.html">Recovery Procedures</a> 79 </span> 80 </dt> 81 <dd> 82 <dl> 83 <dt> 84 <span class="sect2"> 85 <a href="recovery.html#normalrecovery">Normal Recovery</a> 86 </span> 87 </dt> 88 <dt> 89 <span class="sect2"> 90 <a href="recovery.html#catastrophicrecovery">Catastrophic Recovery</a> 91 </span> 92 </dt> 93 </dl> 94 </dd> 95 <dt> 96 <span class="sect1"> 97 <a href="architectrecovery.html">Designing Your Application for Recovery</a> 98 </span> 99 </dt> 100 <dd> 101 <dl> 102 <dt> 103 <span class="sect2"> 104 <a href="architectrecovery.html#multithreadrecovery">Recovery for Multi-Threaded Applications</a> 105 </span> 106 </dt> 107 <dt> 108 <span class="sect2"> 109 <a href="architectrecovery.html#multiprocessrecovery">Recovery in Multi-Process Applications</a> 110 </span> 111 </dt> 112 </dl> 113 </dd> 114 <dt> 115 <span class="sect1"> 116 <a href="hotfailover.html">Using Hot Failovers</a> 117 </span> 118 </dt> 119 <dt> 120 <span class="sect1"> 121 <a href="logfileremoval.html">Removing Log Files</a> 122 </span> 123 </dt> 124 <dt> 125 <span class="sect1"> 126 <a href="logconfig.html">Configuring the Logging Subsystem</a> 127 </span> 128 </dt> 129 <dd> 130 <dl> 131 <dt> 132 <span class="sect2"> 133 <a href="logconfig.html#logfilesize">Setting the Log File Size</a> 134 </span> 135 </dt> 136 <dt> 137 <span class="sect2"> 138 <a href="logconfig.html#logregionsize">Configuring the Logging Region Size</a> 139 </span> 140 </dt> 141 <dt> 142 <span class="sect2"> 143 <a href="logconfig.html#inmemorylogging">Configuring In-Memory Logging</a> 144 </span> 145 </dt> 146 <dt> 147 <span class="sect2"> 148 <a href="logconfig.html#logbuffer">Setting the In-Memory Log Buffer Size</a> 149 </span> 150 </dt> 151 </dl> 152 </dd> 153 </dl> 154 </div> 155 <p> 156 DB is capable of storing several types of files on disk: 157 </p> 158 <div class="itemizedlist"> 159 <ul type="disc"> 160 <li> 161 <p> 162 Data files, which contain the actual data in your database. 163 </p> 164 </li> 165 <li> 166 <p> 167 Log files, which contain information required to recover your 168 database in the event of a system or application failure. 169 </p> 170 </li> 171 <li> 172 <p> 173 Region files, which contain information necessary for the 174 overall operation of your application. 175 </p> 176 </li> 177 <li> 178 <p> 179 Temporary files, which are created only under certain special circumstances. These files never need to 180 be backed up or otherwise managed and so they are not a consideration for the topics described in this 181 chapter. See <a href="enabletxn.html#security">Security Considerations</a> 182 for more information on temporary files. 183 </p> 184 </li> 185 </ul> 186 </div> 187 <p> 188 Of these, you must manage your data and log files by ensuring that they 189 are backed up. You should also pay attention to the amount of disk space 190 your log files are consuming, and periodically remove any unneeded 191 files. Finally, you can optionally tune your logging subsystem to best 192 suit your application's needs and requirements. 193 These topics are discussed in this chapter. 194 </p> 195 <div class="sect1" lang="en" xml:lang="en"> 196 <div class="titlepage"> 197 <div> 198 <div> 199 <h2 class="title" style="clear: both"><a id="checkpoints"></a>Checkpoints</h2> 200 </div> 201 </div> 202 <div></div> 203 </div> 204 <p> 205 Before we can discuss DB file management, we need to 206 describe checkpoints. When databases are modified (that is, a 207 transaction is committed), the modifications are recorded in 208 DB's logs, but they are <span class="emphasis"><em>not</em></span> 209 necessarily reflected in the actual database files on disk. 210 </p> 211 <p> 212 This means that as time goes on, increasingly 213 more data is contained in your log files that is not 214 contained in your data files. As a result, you must keep more 215 log files around than you might actually need. Also, any 216 recovery run from your log files will take increasingly longer 217 amounts of time, because there is more data in the log files 218 that must be reflected back into the data files during the 219 recovery process. 220 </p> 221 <p> 222 You can reduce these problems by periodically 223 running a checkpoint against your environment. The checkpoint: 224 </p> 225 <div class="itemizedlist"> 226 <ul type="disc"> 227 <li> 228 <p> 229 Flushes dirty pages from the in-memory cache. This means that data modifications found in your 230 in-memory cache are written to the database files on disk. Note that a checkpoint also causes data 231 dirtied by an uncommitted transaction to also be written to your database files on disk. In this latter 232 case, DB's normal recovery is used to remove any such modifications that were subsequently 233 abandoned by your application using a transaction abort. 234 </p> 235 <p> 236 Normal recovery is describe in <a href="recovery.html">Recovery Procedures</a>. 237 </p> 238 </li> 239 <li> 240 <p> 241 Writes a checkpoint record. 242 </p> 243 </li> 244 <li> 245 <p> 246 Flushes the log. This causes all log data that has not yet been written to disk to be written. 247 </p> 248 </li> 249 <li> 250 <p> 251 Writes a list of open databases. 252 </p> 253 </li> 254 </ul> 255 </div> 256 <p> 257 There are several ways to run a checkpoint. One way is to use 258 the <span><b class="command">db_checkpoint</b></span> command line utility. (Note, however, that this command line utility 259 cannot be used if your environment was opened using 260 <span> 261 <tt class="literal">DB_PRIVATE</tt>.) 262 </span> 263 264 </p> 265 <p> 266 You can also run a thread that periodically checkpoints your 267 environment for you by calling the 268 <tt class="methodname">DB_ENV->txn_checkpoint()</tt> 269 270 271 method. 272 </p> 273 <p> 274 Note that you can prevent a checkpoint from occurring unless more 275 than a specified amount of log data has been written since the 276 last checkpoint. You can also prevent the checkpoint from 277 running unless more than a specified amount of time has 278 occurred since the last checkpoint. These conditions are 279 particularly interesting if you have multiple threads 280 <span>or processes</span> 281 running checkpoints. 282 </p> 283 <p> 284 For configuration information, see the 285 <a href="http://www.oracle.com/technology/documentation/berkeley-db/db/api_c/txn_checkpoint.html" target="_top"> 286 DB_ENV->txn_checkpoint() API reference page. 287 </a> 288 289 290 </p> 291 <p> 292 Note that running checkpoints can be quite expensive. DB must 293 flush every dirty page to the backing database files. On the 294 other hand, if you do not run checkpoints often enough, your 295 recovery time can be unnecessarily long and you may be using more 296 disk space than you really need. Also, you cannot remove log files 297 until a checkpoint is run. Therefore, deciding how frequently 298 to run a checkpoint is one of the most 299 common tuning activity for DB applications. 300 </p> 301 <p> 302 For example, to run a checkpoint from a separate thread of control: 303 </p> 304 <pre class="programlisting">#include <stdio.h> 305#include <pthread.h> 306#include <stdlib.h> 307#include "db.h" 308 309void *checkpoint_thread(void *); 310 311int 312main(void) 313{ 314 int ret; 315 u_int32_t env_flags; 316 DB_ENV *envp; 317 const char *db_home_dir = "/tmp/myEnvironment"; 318 pthread_t ptid; 319 320 envp = NULL; 321 322 /* Open the environment */ 323 ret = db_env_create(&envp, 0); 324 if (ret != 0) { 325 fprintf(stderr, "Error creating environment handle: %s\n", 326 db_strerror(ret)); 327 return (EXIT_FAILURE); 328 } 329 330 env_flags = DB_CREATE | /* If the environment does not 331 * exist, create it. */ 332 DB_INIT_LOCK | /* Initialize locking */ 333 DB_INIT_LOG | /* Initialize logging */ 334 DB_INIT_MPOOL | /* Initialize the cache */ 335 DB_THREAD | /* Free-thread the env handle. */ 336 DB_INIT_TXN; /* Initialize transactions */ 337 338 /* Open the environment. */ 339 ret = envp->open(envp, db_home_dir, env_flags, 0); 340 if (ret != 0) { 341 fprintf(stderr, "Error opening environment: %s\n", 342 db_strerror(ret)); 343 goto err; 344 } 345 346 347 /* Start a checkpoint thread. */ 348 if ((ret = pthread_create( 349 &ptid, NULL, checkpoint_thread, (void *)envp)) != 0) { 350 fprintf(stderr, 351 "txnapp: failed spawning checkpoint thread: %s\n", 352 strerror(ret)); 353 goto err; 354 } 355 356 /* 357 * All other threads and application shutdown code 358 * omitted for brevity. 359 */ 360 361 ... 362} 363 364 365void * 366checkpoint_thread(void *arg) { 367 DB_ENV *dbenv; 368 int ret; 369 370 dbenv = arg; 371 372 /* Checkpoint once a minute. */ 373 for (;; sleep(60)) 374 if ((ret = dbenv->txn_checkpoint(dbenv, 0, 0, 0)) != 0) { 375 dbenv->err(dbenv, ret, "checkpoint thread"); 376 exit (1); 377 } 378 379 /* NOTREACHED */ 380} </pre> 381 </div> 382 </div> 383 <div class="navfooter"> 384 <hr /> 385 <table width="100%" summary="Navigation footer"> 386 <tr> 387 <td width="40%" align="left"><a accesskey="p" href="reversesplit.html">Prev</a> </td> 388 <td width="20%" align="center"> 389 <a accesskey="u" href="index.html">Up</a> 390 </td> 391 <td width="40%" align="right"> <a accesskey="n" href="backuprestore.html">Next</a></td> 392 </tr> 393 <tr> 394 <td width="40%" align="left" valign="top">Reverse BTree Splits </td> 395 <td width="20%" align="center"> 396 <a accesskey="h" href="index.html">Home</a> 397 </td> 398 <td width="40%" align="right" valign="top"> Backup Procedures</td> 399 </tr> 400 </table> 401 </div> 402 </body> 403</html> 404