Permalink
Fetching contributors…
Cannot retrieve contributors at this time
1623 lines (1288 sloc) 57.1 KB
// d_migrate.cpp
/**
* Copyright (C) 2008 10gen Inc.
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License, version 3,
* as published by the Free Software Foundation.
*
* 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 Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
/**
these are commands that live in mongod
mostly around shard management and checking
*/
#include "pch.h"
#include <map>
#include <string>
#include <algorithm>
#include "../db/commands.h"
#include "../db/jsobj.h"
#include "../db/dbmessage.h"
#include "../db/query.h"
#include "../db/cmdline.h"
#include "../db/queryoptimizer.h"
#include "../db/btree.h"
#include "../db/repl_block.h"
#include "../db/dur.h"
#include "../client/connpool.h"
#include "../client/distlock.h"
#include "../util/queue.h"
#include "../util/unittest.h"
#include "../util/processinfo.h"
#include "../util/ramlog.h"
#include "shard.h"
#include "d_logic.h"
#include "config.h"
#include "chunk.h"
using namespace std;
namespace mongo {
Tee* migrateLog = new RamLog( "migrate" );
class MoveTimingHelper {
public:
MoveTimingHelper( const string& where , const string& ns , BSONObj min , BSONObj max , int total )
: _where( where ) , _ns( ns ) , _next( 0 ) , _total( total ) {
_nextNote = 0;
_b.append( "min" , min );
_b.append( "max" , max );
}
~MoveTimingHelper() {
// even if logChange doesn't throw, bson does
// sigh
try {
if ( _next != _total ) {
note( "aborted" );
}
configServer.logChange( (string)"moveChunk." + _where , _ns, _b.obj() );
}
catch ( const std::exception& e ) {
warning() << "couldn't record timing for moveChunk '" << _where << "': " << e.what() << migrateLog;
}
}
void done( int step ) {
assert( step == ++_next );
assert( step <= _total );
stringstream ss;
ss << "step" << step;
string s = ss.str();
CurOp * op = cc().curop();
if ( op )
op->setMessage( s.c_str() );
else
warning() << "op is null in MoveTimingHelper::done" << migrateLog;
_b.appendNumber( s , _t.millis() );
_t.reset();
#if 0
// debugging for memory leak?
ProcessInfo pi;
ss << " v:" << pi.getVirtualMemorySize()
<< " r:" << pi.getResidentSize();
log() << ss.str() << migrateLog;
#endif
}
void note( const string& s ) {
string field = "note";
if ( _nextNote > 0 ) {
StringBuilder buf;
buf << "note" << _nextNote;
field = buf.str();
}
_nextNote++;
_b.append( field , s );
}
private:
Timer _t;
string _where;
string _ns;
int _next;
int _total; // expected # of steps
int _nextNote;
BSONObjBuilder _b;
};
struct OldDataCleanup {
static AtomicUInt _numThreads; // how many threads are doing async cleanusp
string ns;
BSONObj min;
BSONObj max;
set<CursorId> initial;
OldDataCleanup(){
_numThreads++;
}
OldDataCleanup( const OldDataCleanup& other ) {
ns = other.ns;
min = other.min.getOwned();
max = other.max.getOwned();
initial = other.initial;
_numThreads++;
}
~OldDataCleanup(){
_numThreads--;
}
void doRemove() {
ShardForceVersionOkModeBlock sf;
writelock lk(ns);
RemoveSaver rs("moveChunk",ns,"post-cleanup");
long long num = Helpers::removeRange( ns , min , max , true , false , cmdLine.moveParanoia ? &rs : 0 );
log() << "moveChunk deleted: " << num << migrateLog;
}
};
AtomicUInt OldDataCleanup::_numThreads = 0;
static const char * const cleanUpThreadName = "cleanupOldData";
void _cleanupOldData( OldDataCleanup cleanup ) {
Client::initThread( cleanUpThreadName );
log() << " (start) waiting to cleanup " << cleanup.ns << " from " << cleanup.min << " -> " << cleanup.max << " # cursors:" << cleanup.initial.size() << migrateLog;
int loops = 0;
Timer t;
while ( t.seconds() < 900 ) { // 15 minutes
assert( dbMutex.getState() == 0 );
sleepmillis( 20 );
set<CursorId> now;
ClientCursor::find( cleanup.ns , now );
set<CursorId> left;
for ( set<CursorId>::iterator i=cleanup.initial.begin(); i!=cleanup.initial.end(); ++i ) {
CursorId id = *i;
if ( now.count(id) )
left.insert( id );
}
if ( left.size() == 0 )
break;
cleanup.initial = left;
if ( ( loops++ % 200 ) == 0 ) {
log() << " (looping " << loops << ") waiting to cleanup " << cleanup.ns << " from " << cleanup.min << " -> " << cleanup.max << " # cursors:" << cleanup.initial.size() << migrateLog;
stringstream ss;
for ( set<CursorId>::iterator i=cleanup.initial.begin(); i!=cleanup.initial.end(); ++i ) {
CursorId id = *i;
ss << id << " ";
}
log() << " cursors: " << ss.str() << migrateLog;
}
}
cleanup.doRemove();
cc().shutdown();
}
void cleanupOldData( OldDataCleanup cleanup ) {
try {
_cleanupOldData( cleanup );
}
catch ( std::exception& e ) {
log() << " error cleaning old data:" << e.what() << migrateLog;
}
catch ( ... ) {
log() << " unknown error cleaning old data" << migrateLog;
}
}
class ChunkCommandHelper : public Command {
public:
ChunkCommandHelper( const char * name )
: Command( name ) {
}
virtual void help( stringstream& help ) const {
help << "internal - should not be called directly" << migrateLog;
}
virtual bool slaveOk() const { return false; }
virtual bool adminOnly() const { return true; }
virtual LockType locktype() const { return NONE; }
};
bool isInRange( const BSONObj& obj , const BSONObj& min , const BSONObj& max ) {
BSONObj k = obj.extractFields( min, true );
return k.woCompare( min ) >= 0 && k.woCompare( max ) < 0;
}
class MigrateFromStatus {
public:
MigrateFromStatus() : _m("MigrateFromStatus") {
_active = false;
_inCriticalSection = false;
_memoryUsed = 0;
}
void start( string ns , const BSONObj& min , const BSONObj& max ) {
scoped_lock l(_m); // reads and writes _active
assert( ! _active );
assert( ! min.isEmpty() );
assert( ! max.isEmpty() );
assert( ns.size() );
_ns = ns;
_min = min;
_max = max;
assert( _cloneLocs.size() == 0 );
assert( _deleted.size() == 0 );
assert( _reload.size() == 0 );
assert( _memoryUsed == 0 );
_active = true;
}
void done() {
readlock lk( _ns );
{
scoped_spinlock lk( _trackerLocks );
_deleted.clear();
_reload.clear();
_cloneLocs.clear();
}
_memoryUsed = 0;
scoped_lock l(_m);
_active = false;
_inCriticalSection = false;
}
void logOp( const char * opstr , const char * ns , const BSONObj& obj , BSONObj * patt ) {
if ( ! _getActive() )
return;
if ( _ns != ns )
return;
// no need to log if this is not an insertion, an update, or an actual deletion
// note: opstr 'db' isn't a deletion but a mention that a database exists (for replication
// machinery mostly)
char op = opstr[0];
if ( op == 'n' || op =='c' || ( op == 'd' && opstr[1] == 'b' ) )
return;
BSONElement ide;
if ( patt )
ide = patt->getField( "_id" );
else
ide = obj["_id"];
if ( ide.eoo() ) {
warning() << "logOpForSharding got mod with no _id, ignoring obj: " << obj << migrateLog;
return;
}
BSONObj it;
switch ( opstr[0] ) {
case 'd': {
if ( getThreadName() == cleanUpThreadName ) {
// we don't want to xfer things we're cleaning
// as then they'll be deleted on TO
// which is bad
return;
}
// can't filter deletes :(
_deleted.push_back( ide.wrap() );
_memoryUsed += ide.size() + 5;
return;
}
case 'i':
it = obj;
break;
case 'u':
if ( ! Helpers::findById( cc() , _ns.c_str() , ide.wrap() , it ) ) {
warning() << "logOpForSharding couldn't find: " << ide << " even though should have" << migrateLog;
return;
}
break;
}
if ( ! isInRange( it , _min , _max ) )
return;
_reload.push_back( ide.wrap() );
_memoryUsed += ide.size() + 5;
}
void xfer( list<BSONObj> * l , BSONObjBuilder& b , const char * name , long long& size , bool explode ) {
const long long maxSize = 1024 * 1024;
if ( l->size() == 0 || size > maxSize )
return;
BSONArrayBuilder arr(b.subarrayStart(name));
list<BSONObj>::iterator i = l->begin();
while ( i != l->end() && size < maxSize ) {
BSONObj t = *i;
if ( explode ) {
BSONObj it;
if ( Helpers::findById( cc() , _ns.c_str() , t, it ) ) {
arr.append( it );
size += it.objsize();
}
}
else {
arr.append( t );
}
i = l->erase( i );
size += t.objsize();
}
arr.done();
}
/**
* called from the dest of a migrate
* transfers mods from src to dest
*/
bool transferMods( string& errmsg , BSONObjBuilder& b ) {
if ( ! _getActive() ) {
errmsg = "no active migration!";
return false;
}
long long size = 0;
{
readlock rl( _ns );
Client::Context cx( _ns );
xfer( &_deleted , b , "deleted" , size , false );
xfer( &_reload , b , "reload" , size , true );
}
b.append( "size" , size );
return true;
}
/**
* Get the disklocs that belong to the chunk migrated and sort them in _cloneLocs (to avoid seeking disk later)
*
* @param maxChunkSize number of bytes beyond which a chunk's base data (no indices) is considered too large to move
* @param errmsg filled with textual description of error if this call return false
* @return false if approximate chunk size is too big to move or true otherwise
*/
bool storeCurrentLocs( long long maxChunkSize , string& errmsg , BSONObjBuilder& result ) {
readlock l( _ns );
Client::Context ctx( _ns );
NamespaceDetails *d = nsdetails( _ns.c_str() );
if ( ! d ) {
errmsg = "ns not found, should be impossible";
return false;
}
BSONObj keyPattern;
// the copies are needed because the indexDetailsForRange destroys the input
BSONObj min = _min.copy();
BSONObj max = _max.copy();
IndexDetails *idx = indexDetailsForRange( _ns.c_str() , errmsg , min , max , keyPattern );
if ( idx == NULL ) {
errmsg = "can't find index in storeCurrentLocs";
return false;
}
auto_ptr<ClientCursor> cc( new ClientCursor( QueryOption_NoCursorTimeout ,
shared_ptr<Cursor>( BtreeCursor::make( d , d->idxNo(*idx) , *idx , min , max , false , 1 ) ) ,
_ns ) );
// use the average object size to estimate how many objects a full chunk would carry
// do that while traversing the chunk's range using the sharding index, below
// there's a fair amout of slack before we determine a chunk is too large because object sizes will vary
unsigned long long maxRecsWhenFull;
long long avgRecSize;
const long long totalRecs = d->stats.nrecords;
if ( totalRecs > 0 ) {
avgRecSize = d->stats.datasize / totalRecs;
maxRecsWhenFull = maxChunkSize / avgRecSize;
maxRecsWhenFull = 130 * maxRecsWhenFull / 100; // slack
}
else {
avgRecSize = 0;
maxRecsWhenFull = numeric_limits<long long>::max();
}
scoped_spinlock lk( _trackerLocks );
// do a full traversal of the chunk and don't stop even if we think it is a large chunk
// we want the number of records to better report, in that case
bool isLargeChunk = false;
unsigned long long recCount = 0;;
while ( cc->ok() ) {
DiskLoc dl = cc->currLoc();
if ( ! isLargeChunk ) {
_cloneLocs.insert( dl );
}
cc->advance();
// we can afford to yield here because any change to the base data that we might miss is already being
// queued and will be migrated in the 'transferMods' stage
if ( ! cc->yieldSometimes() ) {
cc.release();
break;
}
if ( ++recCount > maxRecsWhenFull ) {
isLargeChunk = true;
}
}
if ( isLargeChunk ) {
warning() << "can't move chunk of size (aprox) " << recCount * avgRecSize
<< " because maximum size allowed to move is " << maxChunkSize
<< " ns: " << _ns << " " << _min << " -> " << _max
<< migrateLog;
result.appendBool( "chunkTooBig" , true );
result.appendNumber( "estimatedChunkSize" , (long long)(recCount * avgRecSize) );
errmsg = "chunk too big to move";
return false;
}
log() << "moveChunk number of documents: " << _cloneLocs.size() << migrateLog;
return true;
}
bool clone( string& errmsg , BSONObjBuilder& result ) {
if ( ! _getActive() ) {
errmsg = "not active";
return false;
}
ElapsedTracker tracker (128, 10); // same as ClientCursor::_yieldSometimesTracker
int allocSize;
{
readlock l(_ns);
Client::Context ctx( _ns );
NamespaceDetails *d = nsdetails( _ns.c_str() );
assert( d );
scoped_spinlock lk( _trackerLocks );
allocSize = std::min(BSONObjMaxUserSize, (int)((12 + d->averageObjectSize()) * _cloneLocs.size()));
}
BSONArrayBuilder a (allocSize);
while ( 1 ) {
bool filledBuffer = false;
readlock l( _ns );
Client::Context ctx( _ns );
scoped_spinlock lk( _trackerLocks );
set<DiskLoc>::iterator i = _cloneLocs.begin();
for ( ; i!=_cloneLocs.end(); ++i ) {
if (tracker.ping()) // should I yield?
break;
DiskLoc dl = *i;
BSONObj o = dl.obj();
// use the builder size instead of accumulating 'o's size so that we take into consideration
// the overhead of BSONArray indices
if ( a.len() + o.objsize() + 1024 > BSONObjMaxUserSize ) {
filledBuffer = true; // break out of outer while loop
break;
}
a.append( o );
}
_cloneLocs.erase( _cloneLocs.begin() , i );
if ( _cloneLocs.empty() || filledBuffer )
break;
}
result.appendArray( "objects" , a.arr() );
return true;
}
void aboutToDelete( const Database* db , const DiskLoc& dl ) {
dbMutex.assertWriteLocked();
if ( ! _getActive() )
return;
if ( ! db->ownsNS( _ns ) )
return;
// not needed right now
// but trying to prevent a future bug
scoped_spinlock lk( _trackerLocks );
_cloneLocs.erase( dl );
}
long long mbUsed() const { return _memoryUsed / ( 1024 * 1024 ); }
bool getInCriticalSection() const { scoped_lock l(_m); return _inCriticalSection; }
void setInCriticalSection( bool b ) { scoped_lock l(_m); _inCriticalSection = b; }
bool isActive() const { return _getActive(); }
private:
mutable mongo::mutex _m; // protect _inCriticalSection and _active
bool _inCriticalSection;
bool _active;
string _ns;
BSONObj _min;
BSONObj _max;
// we need the lock in case there is a malicious _migrateClone for example
// even though it shouldn't be needed under normal operation
SpinLock _trackerLocks;
// disk locs yet to be transferred from here to the other side
// no locking needed because built initially by 1 thread in a read lock
// emptied by 1 thread in a read lock
// updates applied by 1 thread in a write lock
set<DiskLoc> _cloneLocs;
list<BSONObj> _reload; // objects that were modified that must be recloned
list<BSONObj> _deleted; // objects deleted during clone that should be deleted later
long long _memoryUsed; // bytes in _reload + _deleted
bool _getActive() const { scoped_lock l(_m); return _active; }
void _setActive( bool b ) { scoped_lock l(_m); _active = b; }
} migrateFromStatus;
struct MigrateStatusHolder {
MigrateStatusHolder( string ns , const BSONObj& min , const BSONObj& max ) {
migrateFromStatus.start( ns , min , max );
}
~MigrateStatusHolder() {
migrateFromStatus.done();
}
};
void logOpForSharding( const char * opstr , const char * ns , const BSONObj& obj , BSONObj * patt ) {
migrateFromStatus.logOp( opstr , ns , obj , patt );
}
void aboutToDeleteForSharding( const Database* db , const DiskLoc& dl ) {
migrateFromStatus.aboutToDelete( db , dl );
}
class TransferModsCommand : public ChunkCommandHelper {
public:
TransferModsCommand() : ChunkCommandHelper( "_transferMods" ) {}
bool run(const string& , BSONObj& cmdObj, string& errmsg, BSONObjBuilder& result, bool) {
return migrateFromStatus.transferMods( errmsg, result );
}
} transferModsCommand;
class InitialCloneCommand : public ChunkCommandHelper {
public:
InitialCloneCommand() : ChunkCommandHelper( "_migrateClone" ) {}
bool run(const string& , BSONObj& cmdObj, string& errmsg, BSONObjBuilder& result, bool) {
return migrateFromStatus.clone( errmsg, result );
}
} initialCloneCommand;
/**
* this is the main entry for moveChunk
* called to initial a move
* usually by a mongos
* this is called on the "from" side
*/
class MoveChunkCommand : public Command {
public:
MoveChunkCommand() : Command( "moveChunk" ) {}
virtual void help( stringstream& help ) const {
help << "should not be calling this directly" << migrateLog;
}
virtual bool slaveOk() const { return false; }
virtual bool adminOnly() const { return true; }
virtual LockType locktype() const { return NONE; }
bool run(const string& , BSONObj& cmdObj, string& errmsg, BSONObjBuilder& result, bool) {
// 1. parse options
// 2. make sure my view is complete and lock
// 3. start migrate
// in a read lock, get all DiskLoc and sort so we can do as little seeking as possible
// tell to start transferring
// 4. pause till migrate caught up
// 5. LOCK
// a) update my config, essentially locking
// b) finish migrate
// c) update config server
// d) logChange to config server
// 6. wait for all current cursors to expire
// 7. remove data locally
// -------------------------------
// 1.
string ns = cmdObj.firstElement().str();
string to = cmdObj["to"].str();
string from = cmdObj["from"].str(); // my public address, a tad redundant, but safe
BSONObj min = cmdObj["min"].Obj();
BSONObj max = cmdObj["max"].Obj();
BSONElement shardId = cmdObj["shardId"];
BSONElement maxSizeElem = cmdObj["maxChunkSizeBytes"];
if ( ns.empty() ) {
errmsg = "need to specify namespace in command";
return false;
}
if ( to.empty() ) {
errmsg = "need to specify server to move chunk to";
return false;
}
if ( from.empty() ) {
errmsg = "need to specify server to move chunk from";
return false;
}
if ( min.isEmpty() ) {
errmsg = "need to specify a min";
return false;
}
if ( max.isEmpty() ) {
errmsg = "need to specify a max";
return false;
}
if ( shardId.eoo() ) {
errmsg = "need shardId";
return false;
}
if ( maxSizeElem.eoo() || ! maxSizeElem.isNumber() ) {
errmsg = "need to specify maxChunkSizeBytes";
return false;
}
const long long maxChunkSize = maxSizeElem.numberLong(); // in bytes
if ( ! shardingState.enabled() ) {
if ( cmdObj["configdb"].type() != String ) {
errmsg = "sharding not enabled";
return false;
}
string configdb = cmdObj["configdb"].String();
shardingState.enable( configdb );
configServer.init( configdb );
}
MoveTimingHelper timing( "from" , ns , min , max , 6 /* steps */);
Shard fromShard( from );
Shard toShard( to );
log() << "received moveChunk request: " << cmdObj << migrateLog;
timing.done(1);
// 2.
DistributedLock lockSetup( ConnectionString( shardingState.getConfigServer() , ConnectionString::SYNC ) , ns );
dist_lock_try dlk;
try{
dlk = dist_lock_try( &lockSetup , (string)"migrate-" + min.toString() );
}
catch( LockException& e ){
errmsg = str::stream() << "error locking distributed lock for migration " << "migrate-" << min.toString() << causedBy( e );
return false;
}
if ( ! dlk.got() ) {
errmsg = str::stream() << "the collection metadata could not be locked with lock " << "migrate-" << min.toString();
result.append( "who" , dlk.other() );
return false;
}
BSONObj chunkInfo = BSON("min" << min << "max" << max << "from" << fromShard.getName() << "to" << toShard.getName());
configServer.logChange( "moveChunk.start" , ns , chunkInfo );
ShardChunkVersion maxVersion;
string myOldShard;
{
ScopedDbConnection conn( shardingState.getConfigServer() );
BSONObj x = conn->findOne( ShardNS::chunk , Query( BSON( "ns" << ns ) ).sort( BSON( "lastmod" << -1 ) ) );
maxVersion = x["lastmod"];
BSONObj currChunk = conn->findOne( ShardNS::chunk , shardId.wrap( "_id" ) );
assert( currChunk["shard"].type() );
assert( currChunk["min"].type() );
assert( currChunk["max"].type() );
myOldShard = currChunk["shard"].String();
conn.done();
BSONObj currMin = currChunk["min"].Obj();
BSONObj currMax = currChunk["max"].Obj();
if ( currMin.woCompare( min ) || currMax.woCompare( max ) ) {
errmsg = "boundaries are outdated (likely a split occurred)";
result.append( "currMin" , currMin );
result.append( "currMax" , currMax );
result.append( "requestedMin" , min );
result.append( "requestedMax" , max );
warning() << "aborted moveChunk because" << errmsg << ": " << min << "->" << max
<< " is now " << currMin << "->" << currMax << migrateLog;
return false;
}
if ( myOldShard != fromShard.getName() ) {
errmsg = "location is outdated (likely balance or migrate occurred)";
result.append( "from" , fromShard.getName() );
result.append( "official" , myOldShard );
warning() << "aborted moveChunk because " << errmsg << ": chunk is at " << myOldShard
<< " and not at " << fromShard.getName() << migrateLog;
return false;
}
if ( maxVersion < shardingState.getVersion( ns ) ) {
errmsg = "official version less than mine?";
result.appendTimestamp( "officialVersion" , maxVersion );
result.appendTimestamp( "myVersion" , shardingState.getVersion( ns ) );
warning() << "aborted moveChunk because " << errmsg << ": official " << maxVersion
<< " mine: " << shardingState.getVersion(ns) << migrateLog;
return false;
}
// since this could be the first call that enable sharding we also make sure to have the chunk manager up to date
shardingState.gotShardName( myOldShard );
ShardChunkVersion shardVersion;
shardingState.trySetVersion( ns , shardVersion /* will return updated */ );
log() << "moveChunk request accepted at version " << shardVersion << migrateLog;
}
timing.done(2);
// 3.
MigrateStatusHolder statusHolder( ns , min , max );
{
// this gets a read lock, so we know we have a checkpoint for mods
if ( ! migrateFromStatus.storeCurrentLocs( maxChunkSize , errmsg , result ) )
return false;
ScopedDbConnection connTo( to );
BSONObj res;
bool ok = connTo->runCommand( "admin" ,
BSON( "_recvChunkStart" << ns <<
"from" << from <<
"min" << min <<
"max" << max <<
"configServer" << configServer.modelServer()
) ,
res );
connTo.done();
if ( ! ok ) {
errmsg = "moveChunk failed to engage TO-shard in the data transfer: ";
assert( res["errmsg"].type() );
errmsg += res["errmsg"].String();
result.append( "cause" , res );
return false;
}
}
timing.done( 3 );
// 4.
for ( int i=0; i<86400; i++ ) { // don't want a single chunk move to take more than a day
assert( dbMutex.getState() == 0 );
sleepsecs( 1 );
ScopedDbConnection conn( to );
BSONObj res;
bool ok = conn->runCommand( "admin" , BSON( "_recvChunkStatus" << 1 ) , res );
res = res.getOwned();
conn.done();
log(0) << "moveChunk data transfer progress: " << res << " my mem used: " << migrateFromStatus.mbUsed() << migrateLog;
if ( ! ok || res["state"].String() == "fail" ) {
warning() << "moveChunk error transfering data caused migration abort: " << res << migrateLog;
errmsg = "data transfer error";
result.append( "cause" , res );
return false;
}
if ( res["state"].String() == "steady" )
break;
if ( migrateFromStatus.mbUsed() > (500 * 1024 * 1024) ) {
// this is too much memory for us to use for this
// so we're going to abort the migrate
ScopedDbConnection conn( to );
BSONObj res;
conn->runCommand( "admin" , BSON( "_recvChunkAbort" << 1 ) , res );
res = res.getOwned();
conn.done();
error() << "aborting migrate because too much memory used res: " << res << migrateLog;
errmsg = "aborting migrate because too much memory used";
result.appendBool( "split" , true );
return false;
}
killCurrentOp.checkForInterrupt();
}
timing.done(4);
// 5.
{
// 5.a
// we're under the collection lock here, so no other migrate can change maxVersion or ShardChunkManager state
migrateFromStatus.setInCriticalSection( true );
ShardChunkVersion currVersion = maxVersion;
ShardChunkVersion myVersion = currVersion;
myVersion.incMajor();
{
writelock lk( ns );
assert( myVersion > shardingState.getVersion( ns ) );
// bump the chunks manager's version up and "forget" about the chunk being moved
// this is not the commit point but in practice the state in this shard won't until the commit it done
shardingState.donateChunk( ns , min , max , myVersion );
}
log() << "moveChunk setting version to: " << myVersion << migrateLog;
// 5.b
// we're under the collection lock here, too, so we can undo the chunk donation because no other state change
// could be ongoing
{
BSONObj res;
ScopedDbConnection connTo( to );
bool ok = connTo->runCommand( "admin" ,
BSON( "_recvChunkCommit" << 1 ) ,
res );
connTo.done();
if ( ! ok ) {
{
writelock lk( ns );
// revert the chunk manager back to the state before "forgetting" about the chunk
shardingState.undoDonateChunk( ns , min , max , currVersion );
}
log() << "moveChunk migrate commit not accepted by TO-shard: " << res
<< " resetting shard version to: " << currVersion << migrateLog;
errmsg = "_recvChunkCommit failed!";
result.append( "cause" , res );
return false;
}
log() << "moveChunk migrate commit accepted by TO-shard: " << res << migrateLog;
}
// 5.c
// version at which the next highest lastmod will be set
// if the chunk being moved is the last in the shard, nextVersion is that chunk's lastmod
// otherwise the highest version is from the chunk being bumped on the FROM-shard
ShardChunkVersion nextVersion;
// we want to go only once to the configDB but perhaps change two chunks, the one being migrated and another
// local one (so to bump version for the entire shard)
// we use the 'applyOps' mechanism to group the two updates and make them safer
// TODO pull config update code to a module
BSONObjBuilder cmdBuilder;
BSONArrayBuilder updates( cmdBuilder.subarrayStart( "applyOps" ) );
{
// update for the chunk being moved
BSONObjBuilder op;
op.append( "op" , "u" );
op.appendBool( "b" , false /* no upserting */ );
op.append( "ns" , ShardNS::chunk );
BSONObjBuilder n( op.subobjStart( "o" ) );
n.append( "_id" , Chunk::genID( ns , min ) );
n.appendTimestamp( "lastmod" , myVersion /* same as used on donateChunk */ );
n.append( "ns" , ns );
n.append( "min" , min );
n.append( "max" , max );
n.append( "shard" , toShard.getName() );
n.done();
BSONObjBuilder q( op.subobjStart( "o2" ) );
q.append( "_id" , Chunk::genID( ns , min ) );
q.done();
updates.append( op.obj() );
}
nextVersion = myVersion;
// if we have chunks left on the FROM shard, update the version of one of them as well
// we can figure that out by grabbing the chunkManager installed on 5.a
// TODO expose that manager when installing it
ShardChunkManagerPtr chunkManager = shardingState.getShardChunkManager( ns );
if( chunkManager->getNumChunks() > 0 ) {
// get another chunk on that shard
BSONObj lookupKey;
BSONObj bumpMin, bumpMax;
do {
chunkManager->getNextChunk( lookupKey , &bumpMin , &bumpMax );
lookupKey = bumpMin;
}
while( bumpMin == min );
BSONObjBuilder op;
op.append( "op" , "u" );
op.appendBool( "b" , false );
op.append( "ns" , ShardNS::chunk );
nextVersion.incMinor(); // same as used on donateChunk
BSONObjBuilder n( op.subobjStart( "o" ) );
n.append( "_id" , Chunk::genID( ns , bumpMin ) );
n.appendTimestamp( "lastmod" , nextVersion );
n.append( "ns" , ns );
n.append( "min" , bumpMin );
n.append( "max" , bumpMax );
n.append( "shard" , fromShard.getName() );
n.done();
BSONObjBuilder q( op.subobjStart( "o2" ) );
q.append( "_id" , Chunk::genID( ns , bumpMin ) );
q.done();
updates.append( op.obj() );
log() << "moveChunk updating self version to: " << nextVersion << " through "
<< bumpMin << " -> " << bumpMax << " for collection '" << ns << "'" << migrateLog;
}
else {
log() << "moveChunk moved last chunk out for collection '" << ns << "'" << migrateLog;
}
updates.done();
BSONArrayBuilder preCond( cmdBuilder.subarrayStart( "preCondition" ) );
{
BSONObjBuilder b;
b.append( "ns" , ShardNS::chunk );
b.append( "q" , BSON( "query" << BSON( "ns" << ns ) << "orderby" << BSON( "lastmod" << -1 ) ) );
{
BSONObjBuilder bb( b.subobjStart( "res" ) );
bb.appendTimestamp( "lastmod" , maxVersion );
bb.done();
}
preCond.append( b.obj() );
}
preCond.done();
BSONObj cmd = cmdBuilder.obj();
log(7) << "moveChunk update: " << cmd << migrateLog;
bool ok = false;
BSONObj cmdResult;
try {
ScopedDbConnection conn( shardingState.getConfigServer() );
ok = conn->runCommand( "config" , cmd , cmdResult );
conn.done();
}
catch ( DBException& e ) {
warning() << e << migrateLog;
ok = false;
BSONObjBuilder b;
e.getInfo().append( b );
cmdResult = b.obj();
}
if ( ! ok ) {
// this could be a blip in the connectivity
// wait out a few seconds and check if the commit request made it
//
// if the commit made it to the config, we'll see the chunk in the new shard and there's no action
// if the commit did not make it, currently the only way to fix this state is to bounce the mongod so
// that the old state (before migrating) be brought in
warning() << "moveChunk commit outcome ongoing: " << cmd << " for command :" << cmdResult << migrateLog;
sleepsecs( 10 );
try {
ScopedDbConnection conn( shardingState.getConfigServer() );
// look for the chunk in this shard whose version got bumped
// we assume that if that mod made it to the config, the applyOps was successful
BSONObj doc = conn->findOne( ShardNS::chunk , Query(BSON( "ns" << ns )).sort( BSON("lastmod" << -1)));
ShardChunkVersion checkVersion = doc["lastmod"];
if ( checkVersion == nextVersion ) {
log() << "moveChunk commit confirmed" << migrateLog;
}
else {
error() << "moveChunk commit failed: version is at"
<< checkVersion << " instead of " << nextVersion << migrateLog;
error() << "TERMINATING" << migrateLog;
dbexit( EXIT_SHARDING_ERROR );
}
conn.done();
}
catch ( ... ) {
error() << "moveChunk failed to get confirmation of commit" << migrateLog;
error() << "TERMINATING" << migrateLog;
dbexit( EXIT_SHARDING_ERROR );
}
}
migrateFromStatus.setInCriticalSection( false );
// 5.d
configServer.logChange( "moveChunk.commit" , ns , chunkInfo );
}
migrateFromStatus.done();
timing.done(5);
{
// 6.
OldDataCleanup c;
c.ns = ns;
c.min = min.getOwned();
c.max = max.getOwned();
ClientCursor::find( ns , c.initial );
if ( c.initial.size() ) {
log() << "forking for cleaning up chunk data" << migrateLog;
boost::thread t( boost::bind( &cleanupOldData , c ) );
}
else {
log() << "doing delete inline" << migrateLog;
// 7.
c.doRemove();
}
}
timing.done(6);
return true;
}
} moveChunkCmd;
bool ShardingState::inCriticalMigrateSection() {
return migrateFromStatus.getInCriticalSection();
}
/* -----
below this are the "to" side commands
command to initiate
worker thread
does initial clone
pulls initial change set
keeps pulling
keeps state
command to get state
commend to "commit"
*/
class MigrateStatus {
public:
MigrateStatus() : m_active("MigrateStatus") { active = false; }
void prepare() {
scoped_lock l(m_active); // reading and writing 'active'
assert( ! active );
state = READY;
errmsg = "";
numCloned = 0;
clonedBytes = 0;
numCatchup = 0;
numSteady = 0;
active = true;
}
void go() {
try {
_go();
}
catch ( std::exception& e ) {
state = FAIL;
errmsg = e.what();
error() << "migrate failed: " << e.what() << migrateLog;
}
catch ( ... ) {
state = FAIL;
errmsg = "UNKNOWN ERROR";
error() << "migrate failed with unknown exception" << migrateLog;
}
setActive( false );
}
void _go() {
assert( getActive() );
assert( state == READY );
assert( ! min.isEmpty() );
assert( ! max.isEmpty() );
slaveCount = ( getSlaveCount() / 2 ) + 1;
MoveTimingHelper timing( "to" , ns , min , max , 5 /* steps */ );
ScopedDbConnection conn( from );
conn->getLastError(); // just test connection
{
// 1. copy indexes
auto_ptr<DBClientCursor> indexes = conn->getIndexes( ns );
vector<BSONObj> all;
while ( indexes->more() ) {
all.push_back( indexes->next().getOwned() );
}
writelock lk( ns );
Client::Context ct( ns );
string system_indexes = cc().database()->name + ".system.indexes";
for ( unsigned i=0; i<all.size(); i++ ) {
BSONObj idx = all[i];
theDataFileMgr.insertAndLog( system_indexes.c_str() , idx );
}
timing.done(1);
}
{
// 2. delete any data already in range
writelock lk( ns );
RemoveSaver rs( "moveChunk" , ns , "preCleanup" );
long long num = Helpers::removeRange( ns , min , max , true , false , cmdLine.moveParanoia ? &rs : 0 );
if ( num )
warning() << "moveChunkCmd deleted data already in chunk # objects: " << num << migrateLog;
timing.done(2);
}
{
// 3. initial bulk clone
state = CLONE;
while ( true ) {
BSONObj res;
if ( ! conn->runCommand( "admin" , BSON( "_migrateClone" << 1 ) , res ) ) {
state = FAIL;
errmsg = "_migrateClone failed: ";
errmsg += res.toString();
error() << errmsg << migrateLog;
conn.done();
return;
}
BSONObj arr = res["objects"].Obj();
int thisTime = 0;
BSONObjIterator i( arr );
while( i.more() ) {
BSONObj o = i.next().Obj();
{
writelock lk( ns );
Helpers::upsert( ns , o );
}
thisTime++;
numCloned++;
clonedBytes += o.objsize();
}
if ( thisTime == 0 )
break;
}
timing.done(3);
}
// if running on a replicated system, we'll need to flush the docs we cloned to the secondaries
ReplTime lastOpApplied;
{
// 4. do bulk of mods
state = CATCHUP;
while ( true ) {
BSONObj res;
if ( ! conn->runCommand( "admin" , BSON( "_transferMods" << 1 ) , res ) ) {
state = FAIL;
errmsg = "_transferMods failed: ";
errmsg += res.toString();
error() << "_transferMods failed: " << res << migrateLog;
conn.done();
return;
}
if ( res["size"].number() == 0 )
break;
apply( res , &lastOpApplied );
const int maxIterations = 3600*50;
int i;
for ( i=0;i<maxIterations; i++) {
if ( state == ABORT ) {
timing.note( "aborted" );
return;
}
if ( opReplicatedEnough( lastOpApplied ) )
break;
if ( i > 100 ) {
warning() << "secondaries having hard time keeping up with migrate" << migrateLog;
}
sleepmillis( 20 );
}
if ( i == maxIterations ) {
errmsg = "secondary can't keep up with migrate";
error() << errmsg << migrateLog;
conn.done();
state = FAIL;
return;
}
}
timing.done(4);
}
{
// 5. wait for commit
Timer timeWaitingForCommit;
state = STEADY;
while ( state == STEADY || state == COMMIT_START ) {
BSONObj res;
if ( ! conn->runCommand( "admin" , BSON( "_transferMods" << 1 ) , res ) ) {
log() << "_transferMods failed in STEADY state: " << res << migrateLog;
errmsg = res.toString();
state = FAIL;
conn.done();
return;
}
if ( res["size"].number() > 0 && apply( res , &lastOpApplied ) )
continue;
if ( state == COMMIT_START && flushPendingWrites( lastOpApplied ) )
break;
sleepmillis( 10 );
}
if ( state == ABORT ) {
timing.note( "aborted" );
return;
}
if ( timeWaitingForCommit.seconds() > 86400 ) {
state = FAIL;
errmsg = "timed out waiting for commit";
return;
}
timing.done(5);
}
state = DONE;
conn.done();
}
void status( BSONObjBuilder& b ) {
b.appendBool( "active" , getActive() );
b.append( "ns" , ns );
b.append( "from" , from );
b.append( "min" , min );
b.append( "max" , max );
b.append( "state" , stateString() );
if ( state == FAIL )
b.append( "errmsg" , errmsg );
{
BSONObjBuilder bb( b.subobjStart( "counts" ) );
bb.append( "cloned" , numCloned );
bb.append( "clonedBytes" , clonedBytes );
bb.append( "catchup" , numCatchup );
bb.append( "steady" , numSteady );
bb.done();
}
}
bool apply( const BSONObj& xfer , ReplTime* lastOpApplied ) {
ReplTime dummy;
if ( lastOpApplied == NULL ) {
lastOpApplied = &dummy;
}
bool didAnything = false;
if ( xfer["deleted"].isABSONObj() ) {
writelock lk(ns);
Client::Context cx(ns);
RemoveSaver rs( "moveChunk" , ns , "removedDuring" );
BSONObjIterator i( xfer["deleted"].Obj() );
while ( i.more() ) {
BSONObj id = i.next().Obj();
// do not apply deletes if they do not belong to the chunk being migrated
BSONObj fullObj;
if ( Helpers::findById( cc() , ns.c_str() , id, fullObj ) ) {
if ( ! isInRange( fullObj , min , max ) ) {
log() << "not applying out of range deletion: " << fullObj << migrateLog;
continue;
}
}
Helpers::removeRange( ns , id , id, false , true , cmdLine.moveParanoia ? &rs : 0 );
*lastOpApplied = cx.getClient()->getLastOp();
didAnything = true;
}
}
if ( xfer["reload"].isABSONObj() ) {
writelock lk(ns);
Client::Context cx(ns);
BSONObjIterator i( xfer["reload"].Obj() );
while ( i.more() ) {
BSONObj it = i.next().Obj();
Helpers::upsert( ns , it );
*lastOpApplied = cx.getClient()->getLastOp();
didAnything = true;
}
}
return didAnything;
}
bool opReplicatedEnough( const ReplTime& lastOpApplied ) {
// if replication is on, try to force enough secondaries to catch up
// TODO opReplicatedEnough should eventually honor priorities and geo-awareness
// for now, we try to replicate to a sensible number of secondaries
return mongo::opReplicatedEnough( lastOpApplied , slaveCount );
}
bool flushPendingWrites( const ReplTime& lastOpApplied ) {
if ( ! opReplicatedEnough( lastOpApplied ) ) {
warning() << "migrate commit attempt timed out contacting " << slaveCount
<< " slaves for '" << ns << "' " << min << " -> " << max << migrateLog;
return false;
}
log() << "migrate commit succeeded flushing to secondaries for '" << ns << "' " << min << " -> " << max << migrateLog;
{
readlock lk(ns); // commitNow() currently requires it
// if durability is on, force a write to journal
if ( getDur().commitNow() ) {
log() << "migrate commit flushed to journal for '" << ns << "' " << min << " -> " << max << migrateLog;
}
}
return true;
}
string stateString() {
switch ( state ) {
case READY: return "ready";
case CLONE: return "clone";
case CATCHUP: return "catchup";
case STEADY: return "steady";
case COMMIT_START: return "commitStart";
case DONE: return "done";
case FAIL: return "fail";
case ABORT: return "abort";
}
assert(0);
return "";
}
bool startCommit() {
if ( state != STEADY )
return false;
state = COMMIT_START;
// we wait 5 minutes for the commit to succeed before giving up
for ( int i=0; i<5*60*1000; i++ ) {
sleepmillis(1);
if ( state == DONE )
return true;
}
log() << "startCommit never finished!" << migrateLog;
return false;
}
void abort() {
state = ABORT;
errmsg = "aborted";
}
bool getActive() const { scoped_lock l(m_active); return active; }
void setActive( bool b ) { scoped_lock l(m_active); active = b; }
mutable mongo::mutex m_active;
bool active;
string ns;
string from;
BSONObj min;
BSONObj max;
long long numCloned;
long long clonedBytes;
long long numCatchup;
long long numSteady;
int slaveCount;
enum State { READY , CLONE , CATCHUP , STEADY , COMMIT_START , DONE , FAIL , ABORT } state;
string errmsg;
} migrateStatus;
void migrateThread() {
Client::initThread( "migrateThread" );
migrateStatus.go();
cc().shutdown();
}
class RecvChunkStartCommand : public ChunkCommandHelper {
public:
RecvChunkStartCommand() : ChunkCommandHelper( "_recvChunkStart" ) {}
virtual LockType locktype() const { return WRITE; } // this is so don't have to do locking internally
bool run(const string& , BSONObj& cmdObj, string& errmsg, BSONObjBuilder& result, bool) {
if ( migrateStatus.getActive() ) {
errmsg = "migrate already in progress";
return false;
}
if ( OldDataCleanup::_numThreads > 0 ) {
errmsg =
str::stream()
<< "still waiting for a previous migrates data to get cleaned, can't accept new chunks, num threads: "
<< OldDataCleanup::_numThreads;
return false;
}
if ( ! configServer.ok() )
configServer.init( cmdObj["configServer"].String() );
migrateStatus.prepare();
migrateStatus.ns = cmdObj.firstElement().String();
migrateStatus.from = cmdObj["from"].String();
migrateStatus.min = cmdObj["min"].Obj().getOwned();
migrateStatus.max = cmdObj["max"].Obj().getOwned();
boost::thread m( migrateThread );
result.appendBool( "started" , true );
return true;
}
} recvChunkStartCmd;
class RecvChunkStatusCommand : public ChunkCommandHelper {
public:
RecvChunkStatusCommand() : ChunkCommandHelper( "_recvChunkStatus" ) {}
bool run(const string& , BSONObj& cmdObj, string& errmsg, BSONObjBuilder& result, bool) {
migrateStatus.status( result );
return 1;
}
} recvChunkStatusCommand;
class RecvChunkCommitCommand : public ChunkCommandHelper {
public:
RecvChunkCommitCommand() : ChunkCommandHelper( "_recvChunkCommit" ) {}
bool run(const string& , BSONObj& cmdObj, string& errmsg, BSONObjBuilder& result, bool) {
bool ok = migrateStatus.startCommit();
migrateStatus.status( result );
return ok;
}
} recvChunkCommitCommand;
class RecvChunkAbortCommand : public ChunkCommandHelper {
public:
RecvChunkAbortCommand() : ChunkCommandHelper( "_recvChunkAbort" ) {}
bool run(const string& , BSONObj& cmdObj, string& errmsg, BSONObjBuilder& result, bool) {
migrateStatus.abort();
migrateStatus.status( result );
return true;
}
} recvChunkAboortCommand;
class IsInRangeTest : public UnitTest {
public:
void run() {
BSONObj min = BSON( "x" << 1 );
BSONObj max = BSON( "x" << 5 );
assert( ! isInRange( BSON( "x" << 0 ) , min , max ) );
assert( isInRange( BSON( "x" << 1 ) , min , max ) );
assert( isInRange( BSON( "x" << 3 ) , min , max ) );
assert( isInRange( BSON( "x" << 4 ) , min , max ) );
assert( ! isInRange( BSON( "x" << 5 ) , min , max ) );
assert( ! isInRange( BSON( "x" << 6 ) , min , max ) );
log(1) << "isInRangeTest passed" << migrateLog;
}
} isInRangeTest;
}