--- trunk/bin/BackupPC_updatedb 2005/08/31 20:58:39 107 +++ trunk/bin/BackupPC_updatedb 2006/01/29 14:54:17 311 @@ -11,16 +11,26 @@ use Time::HiRes qw/time/; use File::Pid; use POSIX qw/strftime/; +use BackupPC::SearchLib; +use Cwd qw/abs_path/; use constant BPC_FTYPE_DIR => 5; use constant EST_CHUNK => 100000; +# daylight saving time change offset for 1h +my $dst_offset = 60 * 60; + my $debug = 0; $|=1; my $start_t = time(); -my $pidfile = new File::Pid; +my $pid_path = abs_path($0); +$pid_path =~ s/\W+/_/g; + +my $pidfile = new File::Pid({ + file => "/tmp/$pid_path", +}); if (my $pid = $pidfile->running ) { die "$0 already running: $pid\n"; @@ -28,8 +38,8 @@ $pidfile->remove; $pidfile = new File::Pid; } -$pidfile->write; print STDERR "$0 using pid ",$pidfile->pid," file ",$pidfile->file,"\n"; +$pidfile->write; my $t_fmt = '%Y-%m-%d %H:%M:%S'; @@ -41,28 +51,33 @@ my $dsn = $Conf{SearchDSN} || die "Need SearchDSN in config.pl\n"; my $user = $Conf{SearchUser} || ''; -my $index_path = $Conf{HyperEstraierIndex}; -$index_path = $TopDir . '/' . $index_path; -$index_path =~ s#//#/#g; -if ($index_path) { - use HyperEstraier; -} +my $index_node_url = $Conf{HyperEstraierIndex}; my $dbh = DBI->connect($dsn, $user, "", { RaiseError => 1, AutoCommit => 0 }); my %opt; -if ( !getopts("cdm:v:i", \%opt ) ) { +if ( !getopts("cdm:v:ijfq", \%opt ) ) { print STDERR <sync(); - $hest_db->close(); - } - exit(0); -} - -$SIG{'INT'} = \&signal; -$SIG{'QUIT'} = \&signal; +my $hest_node; sub hest_update { my ($host_id, $share_id, $num) = @_; - print curr_time," updating HyperEstraier:"; + my $skip_check = $opt{j} && print STDERR "Skipping check for existing files -- this should be used only with initital import\n"; + + unless (defined($index_node_url)) { + print STDERR "HyperEstraier support not enabled in configuration\n"; + $index_node_url = 0; + return; + } + + print curr_time," updating Hyper Estraier:"; my $t = time(); my $offset = 0; my $added = 0; - print " opening index $index_path"; - $hest_db = HyperEstraier::Database->new(); - $hest_db->open($index_path, $HyperEstraier::Database::DBWRITER | $HyperEstraier::Database::DBCREAT); - - print " increment is " . EST_CHUNK . " files:"; + print " opening index $index_node_url"; + if ($index_node_url) { + $hest_node ||= Search::Estraier::Node->new( + url => $index_node_url, + user => 'admin', + passwd => 'admin', + croak_on_error => 1, + ); + print " via node URL"; + } else { + die "don't know how to use Hyper Estraier Index $index_node_url"; + } my $results = 0; @@ -125,7 +145,7 @@ my $where = ''; my @data; - if ($host_id && $share_id && $num) { + if (defined($host_id) && defined($share_id) && defined($num)) { $where = qq{ WHERE hosts.id = ? AND @@ -164,7 +184,9 @@ if ($results == 0) { print " - no new files\n"; - last; + return; + } else { + print " - $results files: "; } sub fmt_date { @@ -179,17 +201,20 @@ my $fid = $row->{'fid'} || die "no fid?"; my $uri = 'file:///' . $fid; - my $id = $hest_db->uri_to_id($uri); - next unless ($id == -1); + unless ($skip_check) { + my $id = $hest_node->uri_to_id($uri); + next if ($id && $id == -1); + } # create a document object - my $doc = HyperEstraier::Document->new; + my $doc = Search::Estraier::Document->new; # add attributes to the document object $doc->add_attr('@uri', $uri); foreach my $c (@{ $sth->{NAME} }) { - $doc->add_attr($c, $row->{$c}) if ($row->{$c}); + print STDERR "attr $c = $row->{$c}\n" if ($debug > 2); + $doc->add_attr($c, $row->{$c}) if (defined($row->{$c})); } #$doc->add_attr('@cdate', fmt_date($row->{'date'})); @@ -203,20 +228,20 @@ print STDERR $doc->dump_draft,"\n" if ($debug > 1); # register the document object to the database - $hest_db->put_doc($doc, $HyperEstraier::Database::PDCLEAN); + if ($hest_node) { + $hest_node->put_doc($doc); + } else { + die "not supported"; + } $added++; } print " $added"; - $hest_db->sync(); $offset += EST_CHUNK; } while ($results == EST_CHUNK); - print ", close"; - $hest_db->close(); - my $dur = (time() - $t) || 1; printf(" [%.2f/s dur: %s]\n", ( $added / $dur ), @@ -228,11 +253,11 @@ ## update index ## -if (($opt{i} || ($index_path && ! -e $index_path)) && !$opt{c}) { +if ( ( $opt{i} || $opt{j} ) && !$opt{c} ) { # update all - print "force update of HyperEstraier index "; - print "importing existing data" unless (-e $index_path); + print "force update of Hyper Estraier index "; print "by -i flag" if ($opt{i}); + print "by -j flag" if ($opt{j}); print "\n"; hest_update(); } @@ -241,86 +266,127 @@ if ($opt{c}) { sub do_index { my $index = shift || return; - my ($table,$col,$unique) = split(/_/, $index); + my ($table,$col,$unique) = split(/:/, $index); $unique ||= ''; - $index =~ s/,/_/g; + $index =~ s/\W+/_/g; + print "$index on $table($col)" . ( $unique ? "u" : "" ) . " "; $dbh->do(qq{ create $unique index $index on $table($col) }); } print "creating tables...\n"; - - $dbh->do(qq{ + + $dbh->do( qq{ create table hosts ( ID SERIAL PRIMARY KEY, name VARCHAR(30) NOT NULL, IP VARCHAR(15) ); - }); - - $dbh->do(qq{ + create table shares ( ID SERIAL PRIMARY KEY, hostID INTEGER NOT NULL references hosts(id), name VARCHAR(30) NOT NULL, - share VARCHAR(200) NOT NULL, - localpath VARCHAR(200) + share VARCHAR(200) NOT NULL ); - }); - - $dbh->do(qq{ + + create table dvds ( + ID SERIAL PRIMARY KEY, + num INTEGER NOT NULL, + name VARCHAR(255) NOT NULL, + mjesto VARCHAR(255) + ); + create table backups ( + id serial, hostID INTEGER NOT NULL references hosts(id), num INTEGER NOT NULL, date integer NOT NULL, type CHAR(4) not null, shareID integer not null references shares(id), - size integer not null, - PRIMARY KEY(hostID, num, shareID) + size bigint not null, + inc_size bigint not null default -1, + inc_deleted boolean default false, + parts integer not null default 1, + PRIMARY KEY(id) ); - }); - #do_index('backups_hostid,num_unique'); + create table files ( + ID SERIAL, + shareID INTEGER NOT NULL references shares(id), + backupNum INTEGER NOT NULL, + name VARCHAR(255) NOT NULL, + path VARCHAR(255) NOT NULL, + date integer NOT NULL, + type INTEGER NOT NULL, + size bigint NOT NULL, + primary key(id) + ); - $dbh->do(qq{ - create table dvds ( - ID SERIAL PRIMARY KEY, - num INTEGER NOT NULL, - name VARCHAR(255) NOT NULL, - mjesto VARCHAR(255) + create table archive ( + id serial, + dvd_nr int not null, + total_size bigint default -1, + note text, + username varchar(20) not null, + date timestamp default now(), + primary key(id) + ); + + create table archive_backup ( + archive_id int not null references archive(id) on delete cascade, + backup_id int not null references backups(id), + primary key(archive_id, backup_id) ); - }); - $dbh->do(qq{ - create table files ( - ID SERIAL PRIMARY KEY, - shareID INTEGER NOT NULL references shares(id), - backupNum INTEGER NOT NULL, - name VARCHAR(255) NOT NULL, - path VARCHAR(255) NOT NULL, - date integer NOT NULL, - type INTEGER NOT NULL, - size INTEGER NOT NULL, - dvdid INTEGER references dvds(id) + create table archive_burned ( + archive_id int references archive(id), + date timestamp default now(), + part int not null default 1, + copy int not null default 1, + iso_size bigint default -1 + ); + + create table backup_parts ( + id serial, + backup_id int references backups(id), + part_nr int not null check (part_nr > 0), + tar_size bigint not null check (tar_size > 0), + size bigint not null check (size > 0), + md5 text not null, + items int not null check (items > 0), + date timestamp default now(), + primary key(id) ); }); - print "creating indexes:"; + print "creating indexes: "; foreach my $index (qw( - hosts_name - backups_hostID - backups_num - shares_hostID - shares_name - files_shareID - files_path - files_name - files_date - files_size + hosts:name + backups:hostID + backups:num + backups:shareID + shares:hostID + shares:name + files:shareID + files:path + files:name + files:date + files:size + archive:dvd_nr + archive_burned:archive_id + backup_parts:backup_id,part_nr )) { - print " $index"; do_index($index); } + + print " creating sequence: "; + foreach my $seq (qw/dvd_nr/) { + print "$seq "; + $dbh->do( qq{ CREATE SEQUENCE $seq } ); + } + + print "...\n"; $dbh->commit; @@ -364,7 +430,12 @@ $sth->{insert_backups} = $dbh->prepare(qq{ INSERT INTO backups (hostID, num, date, type, shareid, size) -VALUES (?,?,?,?,?,?) +VALUES (?,?,?,?,?,-1) +}); + +$sth->{update_backups_size} = $dbh->prepare(qq{ +UPDATE backups SET size = ? +WHERE hostID = ? and num = ? and date = ? and type =? and shareid = ? }); $sth->{insert_files} = $dbh->prepare(qq{ @@ -373,7 +444,10 @@ VALUES (?,?,?,?,?,?,?) }); -foreach my $host_key (keys %{$hosts}) { +my @hosts = keys %{$hosts}; +my $host_nr = 0; + +foreach my $host_key (@hosts) { my $hostname = $hosts->{$host_key}->{'host'} || die "can't find host for $host_key"; @@ -388,13 +462,19 @@ $hostID = $dbh->last_insert_id(undef,undef,'hosts',undef); } - print "host ".$hosts->{$host_key}->{'host'}.": "; - + $host_nr++; # get backups for a host my @backups = $bpc->BackupInfoRead($hostname); my $incs = scalar @backups; - print "$incs increments\n"; + my $host_header = sprintf("host %s [%d/%d]: %d increments\n", + $hosts->{$host_key}->{'host'}, + $host_nr, + ($#hosts + 1), + $incs + ); + print $host_header unless ($opt{q}); + my $inc_nr = 0; $beenThere = {}; @@ -406,7 +486,7 @@ my $backupNum = $backup->{'num'}; my @backupShares = (); - printf("%-10s %2d/%-2d #%-2d %s %5s/%5s files (date: %s dur: %s)\n", + my $share_header = sprintf("%-10s %2d/%-2d #%-2d %s %5s/%5s files (date: %s dur: %s)\n", $hosts->{$host_key}->{'host'}, $inc_nr, $incs, $backupNum, $backup->{type} || '?', @@ -414,6 +494,7 @@ strftime($t_fmt,localtime($backup->{startTime})), fmt_time($backup->{endTime} - $backup->{startTime}) ); + print $share_header unless ($opt{q}); my $files = BackupPC::View->new($bpc, $hostname, \@backups, 1); foreach my $share ($files->shareList($backupNum)) { @@ -427,22 +508,44 @@ # skip if allready in database! next if ($count > 0); + # dump host and share header for -q + if ($opt{q}) { + if ($host_header) { + print $host_header; + $host_header = undef; + } + print $share_header; + } + # dump some log print curr_time," ", $share; - my ($f, $nf, $d, $nd, $size) = recurseDir($bpc, $hostname, $files, $backupNum, $share, "", $shareID); - $sth->{insert_backups}->execute( $hostID, $backupNum, $backup->{'endTime'}, - $backup->{'type'}, + substr($backup->{'type'},0,4), $shareID, - $size, ); - print " commit"; - $dbh->commit(); + my ($f, $nf, $d, $nd, $size) = recurseDir($bpc, $hostname, $files, $backupNum, $share, "", $shareID); + + eval { + $sth->{update_backups_size}->execute( + $size, + $hostID, + $backupNum, + $backup->{'endTime'}, + substr($backup->{'type'},0,4), + $shareID, + ); + print " commit"; + $dbh->commit(); + }; + if ($@) { + print " rollback"; + $dbh->rollback(); + } my $dur = (time() - $t) || 1; printf(" %d/%d files %d/%d dirs %0.2f MB [%.2f/s dur: %s]\n", @@ -481,14 +584,14 @@ $sth->{insert_share} ||= $dbh->prepare(qq{ INSERT INTO shares - (hostID,name,share,localpath) - VALUES (?,?,?,?) + (hostID,name,share) + VALUES (?,?,?) }); my $drop_down = $hostname . '/' . $share; $drop_down =~ s#//+#/#g; - $sth->{insert_share}->execute($hostID,$share, $drop_down ,undef); + $sth->{insert_share}->execute($hostID,$share, $drop_down); return $dbh->last_insert_id(undef,undef,'shares',undef); } @@ -505,12 +608,12 @@ SELECT 1 FROM files WHERE shareID = ? and path = ? and - date = ? and - size = ? + size = ? and + ( date = ? or date = ? or date = ? ) LIMIT 1 }); - my @param = ($shareID,$path,$date,$size); + my @param = ($shareID,$path,$size,$date, $date-$dst_offset, $date+$dst_offset); $sth->{file_in_db}->execute(@param); my $rows = $sth->{file_in_db}->rows; print STDERR "## found_in_db($shareID,$path,$date,$size) ",( $rows ? '+' : '-' ), join(" ",@param), "\n" if ($debug >= 3); @@ -560,8 +663,29 @@ $filesInBackup->{$path_key}->{'size'} )); + my $key_dst_prev = join(" ", ( + $shareID, + $dir, + $path_key, + $filesInBackup->{$path_key}->{'mtime'} - $dst_offset, + $filesInBackup->{$path_key}->{'size'} + )); + + my $key_dst_next = join(" ", ( + $shareID, + $dir, + $path_key, + $filesInBackup->{$path_key}->{'mtime'} + $dst_offset, + $filesInBackup->{$path_key}->{'size'} + )); + my $found; - if (! defined($beenThere->{$key}) && ! ($found = found_in_db($key, @data)) ) { + if ( + ! defined($beenThere->{$key}) && + ! defined($beenThere->{$key_dst_prev}) && + ! defined($beenThere->{$key_dst_next}) && + ! ($found = found_in_db($key, @data)) + ) { print STDERR "# key: $key [", $beenThere->{$key},"]" if ($debug >= 2); if ($filesInBackup->{$path_key}->{'type'} == BPC_FTYPE_DIR) {