NdbBackup.cpp 10.3 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33
/* Copyright (C) 2003 MySQL AB

   This program is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; either version 2 of the License, or
   (at your option) any later version.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with this program; if not, write to the Free Software
   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA */

#include <signaldata/DumpStateOrd.hpp>
#include <NdbBackup.hpp>
#include <NdbOut.hpp>
#include <NDBT_Output.hpp>
#include <NdbConfig.h>
#include <ndb_version.h>
#include <NDBT.hpp>
#include <NdbSleep.h>
#include <random.h>
#include <NdbTick.h>

#define CHECK(b, m) { int _xx = b; if (!(_xx)) { \
  ndbout << "ERR: "<< m \
           << "   " << "File: " << __FILE__ \
           << " (Line: " << __LINE__ << ")" << "- " << _xx << endl; \
  return NDBT_FAILED; } }

34 35 36 37
#include <ConfigRetriever.hpp>
#include <mgmapi.h>
#include <mgmapi_config_parameters.h>
#include <mgmapi_configuration.hpp>
38 39 40 41 42 43 44 45 46 47 48

int 
NdbBackup::start(unsigned int & _backup_id){

  
  if (!isConnected())
    return -1;

  ndb_mgm_reply reply;
  reply.return_code = 0;

tomas@poseidon.ndb.mysql.com's avatar
tomas@poseidon.ndb.mysql.com committed
49 50
  if (ndb_mgm_start_backup(handle,
			   2, // wait until completed
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67
			   &_backup_id,
			   &reply) == -1) {
    g_err  << "Could not start backup " << endl;
    g_err << "Error: " << reply.message << endl;
    return -1;
  }

  if(reply.return_code != 0){
    g_err  << "PLEASE CHECK CODE NdbBackup.cpp line=" << __LINE__ << endl;
    g_err << "Error: " << reply.message << endl;
    return reply.return_code;
  }
  return 0;
}


const char * 
tomas@poseidon.(none)'s avatar
tomas@poseidon.(none) committed
68
NdbBackup::getBackupDataDirForNode(int _node_id){
69 70 71 72

  /**
   * Fetch configuration from management server
   */
73 74 75 76 77 78 79 80 81
  ndb_mgm_configuration *p;
  if (connect())
    return NULL;

  if ((p = ndb_mgm_get_configuration(handle, 0)) == 0)
  {
    const char * s= ndb_mgm_get_latest_error_msg(handle);
    if(s == 0)
      s = "No error given!";
82
      
83 84 85
    ndbout << "Could not fetch configuration" << endl;
    ndbout << s << endl;
    return NULL;
86 87 88 89 90
  }
  
  /**
   * Setup cluster configuration data
   */
91 92
  ndb_mgm_configuration_iterator iter(* p, CFG_SECTION_NODE);
  if (iter.find(CFG_NODE_ID, _node_id)){
93 94 95
    ndbout << "Invalid configuration fetched, DB missing" << endl;
    return NULL;
  }
joreland@mysql.com's avatar
joreland@mysql.com committed
96 97

  unsigned int type = NODE_TYPE_DB + 1;
98 99
  if(iter.get(CFG_TYPE_OF_SECTION, &type) || type != NODE_TYPE_DB){
    ndbout <<"type = " << type << endl;
100 101 102
    ndbout <<"Invalid configuration fetched, I'm wrong type of node" << endl;
    return NULL;
  }  
103
  
104
  const char * path;
tomas@poseidon.(none)'s avatar
tomas@poseidon.(none) committed
105 106
  if (iter.get(CFG_DB_BACKUP_DATADIR, &path)){
    ndbout << "BackupDataDir not found" << endl;
107 108 109 110 111 112 113 114 115 116 117 118 119 120 121
    return NULL;
  }

  return path;

}

int  
NdbBackup::execRestore(bool _restore_data,
		       bool _restore_meta,
		       int _node_id,
		       unsigned _backup_id){
  const int buf_len = 1000;
  char buf[buf_len];

tomas@poseidon.(none)'s avatar
tomas@poseidon.(none) committed
122
  ndbout << "getBackupDataDir "<< _node_id <<endl;
123

tomas@poseidon.(none)'s avatar
tomas@poseidon.(none) committed
124
  const char* path = getBackupDataDirForNode(_node_id);
125 126 127
  if (path == NULL)
    return -1;  

128
  ndbout << "getHostName "<< _node_id <<endl;
129 130 131 132 133 134 135 136 137
  const char *host;
  if (!getHostName(_node_id, &host)){
    return -1;
  }

  /* 
   * Copy  backup files to local dir
   */ 

138
  BaseString::snprintf(buf, buf_len,
139
	   "scp %s:%s/BACKUP/BACKUP-%d/BACKUP-%d*.%d.* .",
140
	   host, path,
141 142 143
	   _backup_id,
	   _backup_id,
	   _node_id);
144 145 146 147

  ndbout << "buf: "<< buf <<endl;
  int res = system(buf);  
  
148
  ndbout << "scp res: " << res << endl;
149
  
150
  BaseString::snprintf(buf, 255, "%sndb_restore -c \"%s:%d\" -n %d -b %d %s %s .", 
joreland@mysql.com's avatar
joreland@mysql.com committed
151 152 153 154
#if 1
	   "",
#else
	   "valgrind --leak-check=yes -v "
155
#endif
156 157
	   ndb_mgm_get_connected_host(handle),
	   ndb_mgm_get_connected_port(handle),
158 159 160 161 162 163
	   _node_id, 
	   _backup_id,
	   _restore_data?"-r":"",
	   _restore_meta?"-m":"");

  ndbout << "buf: "<< buf <<endl;
164
  res = system(buf);
165

166
  ndbout << "ndb_restore res: " << res << endl;
167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182

  return res;
  
}

int 
NdbBackup::restore(unsigned _backup_id){
  
  if (!isConnected())
    return -1;

  if (getStatus() != 0)
    return -1;

  int res; 

183 184 185 186 187 188
  // restore metadata first and data for first node
  res = execRestore(true, true, ndbNodes[0].node_id, _backup_id);

  // Restore data once for each node
  for(size_t i = 1; i < ndbNodes.size(); i++){
    res = execRestore(true, false, ndbNodes[i].node_id, _backup_id);
189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440
  }
  
  return 0;
}

// Master failure
int
NFDuringBackupM_codes[] = {
  10003,
  10004,
  10005,
  10007,
  10008,
  10009,
  10010,
  10012,
  10013
};

// Slave failure
int
NFDuringBackupS_codes[] = {
  10014,
  10015,
  10016,
  10017,
  10018,
  10020
};

// Master takeover etc...
int
NFDuringBackupSL_codes[] = {
  10001,
  10002,
  10021
};

int 
NdbBackup::NFMaster(NdbRestarter& _restarter){
  const int sz = sizeof(NFDuringBackupM_codes)/sizeof(NFDuringBackupM_codes[0]);
  return NF(_restarter, NFDuringBackupM_codes, sz, true);
}

int 
NdbBackup::NFMasterAsSlave(NdbRestarter& _restarter){
  const int sz = sizeof(NFDuringBackupS_codes)/sizeof(NFDuringBackupS_codes[0]);
  return NF(_restarter, NFDuringBackupS_codes, sz, true);
}

int 
NdbBackup::NFSlave(NdbRestarter& _restarter){
  const int sz = sizeof(NFDuringBackupS_codes)/sizeof(NFDuringBackupS_codes[0]);
  return NF(_restarter, NFDuringBackupS_codes, sz, false);
}

int 
NdbBackup::NF(NdbRestarter& _restarter, int *NFDuringBackup_codes, const int sz, bool onMaster){
  {
    int nodeId = _restarter.getMasterNodeId();

    CHECK(_restarter.restartOneDbNode(nodeId, false, true, true) == 0,
	  "Could not restart node "<< nodeId);
    
    CHECK(_restarter.waitNodesNoStart(&nodeId, 1) == 0,
	  "waitNodesNoStart failed");
    
    CHECK(_restarter.startNodes(&nodeId, 1) == 0,
	  "failed to start node");

    NdbSleep_SecSleep(10);
  }

  CHECK(_restarter.waitClusterStarted() == 0,
	"waitClusterStarted failed");

  int nNodes = _restarter.getNumDbNodes();

  myRandom48Init(NdbTick_CurrentMillisecond());

  for(int i = 0; i<sz; i++){

    int error = NFDuringBackup_codes[i];
    unsigned int backupId;

    const int masterNodeId = _restarter.getMasterNodeId();
    CHECK(masterNodeId > 0, "getMasterNodeId failed");
    int nodeId;

    nodeId = masterNodeId;
    if (!onMaster) {
      int randomId;
      while (nodeId == masterNodeId) {
	randomId = myRandom48(nNodes);
	nodeId = _restarter.getDbNodeId(randomId);
      }
    }

    g_err << "NdbBackup::NF node = " << nodeId 
	   << " error code = " << error << " masterNodeId = "
	   << masterNodeId << endl;


    int val = DumpStateOrd::CmvmiSetRestartOnErrorInsert;
    CHECK(_restarter.dumpStateOneNode(nodeId, &val, 1) == 0,
	  "failed to set RestartOnErrorInsert");
    CHECK(_restarter.insertErrorInNode(nodeId, error) == 0,
	  "failed to set error insert");
   
    g_info << "error inserted"  << endl;

    g_info << "starting backup"  << endl;
    int r = start(backupId);
    g_info << "r = " << r
	   << " (which should fail) started with id = "  << backupId << endl;
    if (r == 0) {
      g_err << "Backup should have failed on error_insertion " << error << endl
	    << "Master = " << masterNodeId << "Node = " << nodeId << endl;
    }

    CHECK(_restarter.waitNodesNoStart(&nodeId, 1) == 0,
	  "waitNodesNoStart failed");

    g_info << "number of nodes running " << _restarter.getNumDbNodes() << endl;

    if (_restarter.getNumDbNodes() != nNodes) {
      g_err << "Failure: cluster not up" << endl;
      return NDBT_FAILED;
    }

    NdbSleep_SecSleep(1);

    g_info << "starting new backup"  << endl;
    CHECK(start(backupId) == 0,
	  "failed to start backup");
    g_info << "(which should succeed) started with id = "  << backupId << endl;

    g_info << "starting node"  << endl;
    CHECK(_restarter.startNodes(&nodeId, 1) == 0,
	  "failed to start node");

    CHECK(_restarter.waitClusterStarted() == 0,
	  "waitClusterStarted failed");
    g_info << "node started"  << endl;

    CHECK(_restarter.insertErrorInNode(nodeId, 10099) == 0,
	  "failed to set error insert");
  }

  return NDBT_OK;
};

int
FailS_codes[] = {
  10023,
  10024,
  10025,
  10026,
  10027,
  10028,
  10029,
  10030,
  10031
};

int
FailM_codes[] = {
  10023,
  10024,
  10025,
  10026,
  10027,
  10028,
  10029,
  10030,
  10031
};

int 
NdbBackup::FailMaster(NdbRestarter& _restarter){
  const int sz = sizeof(FailM_codes)/sizeof(FailM_codes[0]);
  return Fail(_restarter, FailM_codes, sz, true);
}

int 
NdbBackup::FailMasterAsSlave(NdbRestarter& _restarter){
  const int sz = sizeof(FailS_codes)/sizeof(FailS_codes[0]);
  return Fail(_restarter, FailS_codes, sz, true);
}

int 
NdbBackup::FailSlave(NdbRestarter& _restarter){
  const int sz = sizeof(FailS_codes)/sizeof(FailS_codes[0]);
  return Fail(_restarter, FailS_codes, sz, false);
}

int 
NdbBackup::Fail(NdbRestarter& _restarter, int *Fail_codes, const int sz, bool onMaster){

  CHECK(_restarter.waitClusterStarted() == 0,
	"waitClusterStarted failed");

  int nNodes = _restarter.getNumDbNodes();

  myRandom48Init(NdbTick_CurrentMillisecond());

  for(int i = 0; i<sz; i++){
    int error = Fail_codes[i];
    unsigned int backupId;

    const int masterNodeId = _restarter.getMasterNodeId();
    CHECK(masterNodeId > 0, "getMasterNodeId failed");
    int nodeId;

    nodeId = masterNodeId;
    if (!onMaster) {
      int randomId;
      while (nodeId == masterNodeId) {
	randomId = myRandom48(nNodes);
	nodeId = _restarter.getDbNodeId(randomId);
      }
    }

    g_err << "NdbBackup::Fail node = " << nodeId 
	   << " error code = " << error << " masterNodeId = "
	   << masterNodeId << endl;

    CHECK(_restarter.insertErrorInNode(nodeId, error) == 0,
	  "failed to set error insert");
   
    g_info << "error inserted"  << endl;
    g_info << "waiting some before starting backup"  << endl;

    g_info << "starting backup"  << endl;
    int r = start(backupId);
    g_info << "r = " << r
	   << " (which should fail) started with id = "  << backupId << endl;
    if (r == 0) {
      g_err << "Backup should have failed on error_insertion " << error << endl
	    << "Master = " << masterNodeId << "Node = " << nodeId << endl;
    }

    CHECK(_restarter.waitClusterStarted() == 0,
	  "waitClusterStarted failed");

    CHECK(_restarter.insertErrorInNode(nodeId, 10099) == 0,
	  "failed to set error insert");
  }

  return NDBT_OK;
}