Merge branch 'master' into reset-resume-state
This commit is contained in:
Christoph Klaffl
commit
a9e540ebcd
2
INSTALL
2
INSTALL
|
|
@ -8,7 +8,7 @@ default for SSH transport since v1.4.6. Syncoid runs will fail if one of them
|
|||
is not available on either end of the transport.
|
||||
|
||||
On Ubuntu: apt install pv lzop mbuffer
|
||||
On CentOS: yum install lzo pv mbuffer lzop
|
||||
On CentOS: yum install lzo pv mbuffer lzop perl-Data-Dumper
|
||||
On FreeBSD: pkg install pv mbuffer lzop
|
||||
|
||||
FreeBSD notes: FreeBSD may place pv and lzop in somewhere other than
|
||||
|
|
|
|||
34
README.md
34
README.md
|
|
@ -118,6 +118,32 @@ If ZFS supports resumeable send/receive streams on both the source and target th
|
|||
|
||||
As of 1.4.18, syncoid also automatically supports and enables resume of interrupted replication when both source and target support this feature.
|
||||
|
||||
##### Syncoid Dataset Properties
|
||||
|
||||
+ syncoid:sync
|
||||
|
||||
Available values:
|
||||
|
||||
+ `true` (default if unset)
|
||||
|
||||
This dataset will be synchronised to all hosts.
|
||||
|
||||
+ `false`
|
||||
|
||||
This dataset will not be synchronised to any hosts - it will be skipped. This can be useful for preventing certain datasets from being transferred when recursively handling a tree.
|
||||
|
||||
+ `host1,host2,...`
|
||||
|
||||
A comma separated list of hosts. This dataset will only be synchronised by hosts listed in the property.
|
||||
|
||||
_Note_: this check is performed by the host running `syncoid`, thus the local hostname must be present for inclusion during a push operation // the remote hostname must be present for a pull.
|
||||
|
||||
_Note_: this will also prevent syncoid from handling the dataset if given explicitly on the command line.
|
||||
|
||||
_Note_: syncing a child of a no-sync dataset will currently result in a critical error.
|
||||
|
||||
_Note_: empty properties will be handled as if they were unset.
|
||||
|
||||
##### Syncoid Command Line Options
|
||||
|
||||
+ [source]
|
||||
|
|
@ -128,10 +154,18 @@ As of 1.4.18, syncoid also automatically supports and enables resume of interrup
|
|||
|
||||
This is the destination dataset. It can be either local or remote.
|
||||
|
||||
+ --identifier=
|
||||
|
||||
Adds the given identifier to the snapshot name after "syncoid_" prefix and before the hostname. This enables the use case of reliable replication to multiple targets from the same host. The following chars are allowed: a-z, A-Z, 0-9, _, -, : and . .
|
||||
|
||||
+ -r --recursive
|
||||
|
||||
This will also transfer child datasets.
|
||||
|
||||
+ --skip-parent
|
||||
|
||||
This will skip the syncing of the parent dataset. Does nothing without '--recursive' option.
|
||||
|
||||
+ --compress <compression type>
|
||||
|
||||
Currently accepted options: gzip, pigz-fast, pigz-slow, lzo (default) & none. If the selected compression method is unavailable on the source and destination, no compression will be used.
|
||||
|
|
|
|||
7
sanoid
7
sanoid
|
|
@ -976,6 +976,11 @@ sub check_zpool() {
|
|||
## other cases
|
||||
my ($dev, $sta) = /^\s+(\S+)\s+(\S+)/;
|
||||
|
||||
if (!defined($sta)) {
|
||||
# cache and logs are special and don't have a status
|
||||
next;
|
||||
}
|
||||
|
||||
## pool online, not degraded thanks to dead/corrupted disk
|
||||
if ($state eq "OK" && $sta eq "UNAVAIL") {
|
||||
$state="WARNING";
|
||||
|
|
@ -1111,7 +1116,7 @@ sub checklock {
|
|||
# make sure lockfile contains something
|
||||
if ( -z $lockfile) {
|
||||
# zero size lockfile, something is wrong
|
||||
die "ERROR: something is wrong! $lockfile is empty\n";
|
||||
die "ERROR: something is wrong! $lockfile is empty\n";
|
||||
}
|
||||
|
||||
# lockfile exists. read pid and mutex from it. see if it's our pid. if not, see if
|
||||
|
|
|
|||
121
syncoid
121
syncoid
|
|
@ -19,7 +19,7 @@ use Sys::Hostname;
|
|||
my %args = ('sshkey' => '', 'sshport' => '', 'sshcipher' => '', 'sshoption' => [], 'target-bwlimit' => '', 'source-bwlimit' => '');
|
||||
GetOptions(\%args, "no-command-checks", "monitor-version", "compress=s", "dumpsnaps", "recursive|r",
|
||||
"source-bwlimit=s", "target-bwlimit=s", "sshkey=s", "sshport=i", "sshcipher|c=s", "sshoption|o=s@",
|
||||
"debug", "quiet", "no-stream", "no-sync-snap", "no-resume", "exclude=s@") or pod2usage(2);
|
||||
"debug", "quiet", "no-stream", "no-sync-snap", "no-resume", "exclude=s@", "skip-parent", "identifier=s") or pod2usage(2);
|
||||
|
||||
my %compressargs = %{compressargset($args{'compress'} || 'default')}; # Can't be done with GetOptions arg, as default still needs to be set
|
||||
|
||||
|
|
@ -71,6 +71,17 @@ if (length $args{'sshkey'}) {
|
|||
}
|
||||
my $sshoptions = join " ", map { "-o " . $_ } @{$args{'sshoption'}}; # deref required
|
||||
|
||||
my $identifier = "";
|
||||
if (length $args{'identifier'}) {
|
||||
if ($args{'identifier'} !~ /^[a-zA-Z0-9-_:.]+$/) {
|
||||
# invalid extra identifier
|
||||
print("CRITICAL: extra identifier contains invalid chars!\n");
|
||||
pod2usage(2);
|
||||
exit 127;
|
||||
}
|
||||
$identifier = "$args{'identifier'}_";
|
||||
}
|
||||
|
||||
# figure out if source and/or target are remote.
|
||||
$sshcmd = "$sshcmd $args{'sshcipher'} $sshoptions $args{'sshport'} $args{'sshkey'}";
|
||||
if ($debug) { print "DEBUG: SSHCMD: $sshcmd\n"; }
|
||||
|
|
@ -86,6 +97,7 @@ my $targetsudocmd = $targetisroot ? '' : $sudocmd;
|
|||
my %avail = checkcommands();
|
||||
|
||||
my %snaps;
|
||||
my $exitcode = 0;
|
||||
|
||||
## break here to call replication individually so that we ##
|
||||
## can loop across children separately, for recursive ##
|
||||
|
|
@ -116,7 +128,7 @@ if ($targethost ne '') {
|
|||
close FH;
|
||||
}
|
||||
|
||||
exit 0;
|
||||
exit $exitcode;
|
||||
|
||||
##############################################################################
|
||||
##############################################################################
|
||||
|
|
@ -141,6 +153,11 @@ sub getchilddatasets {
|
|||
my @children = <FH>;
|
||||
close FH;
|
||||
|
||||
if (defined $args{'skip-parent'}) {
|
||||
# parent dataset is the first element
|
||||
shift @children;
|
||||
}
|
||||
|
||||
if (defined $args{'exclude'}) {
|
||||
my $excludes = $args{'exclude'};
|
||||
foreach (@$excludes) {
|
||||
|
|
@ -167,9 +184,27 @@ sub syncdataset {
|
|||
|
||||
if ($debug) { print "DEBUG: syncing source $sourcefs to target $targetfs.\n"; }
|
||||
|
||||
my $sync = getzfsvalue($sourcehost,$sourcefs,$sourceisroot,'syncoid:sync');
|
||||
|
||||
if ($sync eq 'true' || $sync eq '-' || $sync eq '') {
|
||||
# empty is handled the same as unset (aka: '-')
|
||||
# definitely sync this dataset - if a host is called 'true' or '-', then you're special
|
||||
} elsif ($sync eq 'false') {
|
||||
if (!$quiet) { print "INFO: Skipping dataset (syncoid:sync=false): $sourcefs...\n"; }
|
||||
return 0;
|
||||
} else {
|
||||
my $hostid = hostname();
|
||||
my @hosts = split(/,/,$sync);
|
||||
if (!(grep $hostid eq $_, @hosts)) {
|
||||
if (!$quiet) { print "INFO: Skipping dataset (syncoid:sync doesn't include $hostid): $sourcefs...\n"; }
|
||||
return 0;
|
||||
}
|
||||
}
|
||||
|
||||
# make sure target is not currently in receive.
|
||||
if (iszfsbusy($targethost,$targetfs,$targetisroot)) {
|
||||
warn "Cannot sync now: $targetfs is already target of a zfs receive process.\n";
|
||||
if ($exitcode < 1) { $exitcode = 1; }
|
||||
return 0;
|
||||
}
|
||||
|
||||
|
|
@ -215,11 +250,16 @@ sub syncdataset {
|
|||
if (!defined $args{'no-sync-snap'}) {
|
||||
# create a new syncoid snapshot on the source filesystem.
|
||||
$newsyncsnap = newsyncsnap($sourcehost,$sourcefs,$sourceisroot);
|
||||
if (!$newsyncsnap) {
|
||||
# we already whined about the error
|
||||
return 0;
|
||||
}
|
||||
} else {
|
||||
# we don't want sync snapshots created, so use the newest snapshot we can find.
|
||||
$newsyncsnap = getnewestsnapshot($sourcehost,$sourcefs,$sourceisroot);
|
||||
if ($newsyncsnap eq 0) {
|
||||
warn "CRITICAL: no snapshots exist on source $sourcefs, and you asked for --no-sync-snap.\n";
|
||||
if ($exitcode < 1) { $exitcode = 1; }
|
||||
return 0;
|
||||
}
|
||||
}
|
||||
|
|
@ -248,6 +288,11 @@ sub syncdataset {
|
|||
}
|
||||
my $oldestsnap = getoldestsnapshot(\%snaps);
|
||||
if (! $oldestsnap) {
|
||||
if (defined ($args{'no-sync-snap'}) ) {
|
||||
# we already whined about the missing snapshots
|
||||
return 0;
|
||||
}
|
||||
|
||||
# getoldestsnapshot() returned false, so use new sync snapshot
|
||||
if ($debug) { print "DEBUG: getoldestsnapshot() returned false, so using $newsyncsnap.\n"; }
|
||||
$oldestsnap = $newsyncsnap;
|
||||
|
|
@ -276,10 +321,14 @@ sub syncdataset {
|
|||
# make sure target is (still) not currently in receive.
|
||||
if (iszfsbusy($targethost,$targetfs,$targetisroot)) {
|
||||
warn "Cannot sync now: $targetfs is already target of a zfs receive process.\n";
|
||||
if ($exitcode < 1) { $exitcode = 1; }
|
||||
return 0;
|
||||
}
|
||||
system($synccmd) == 0
|
||||
or die "CRITICAL ERROR: $synccmd failed: $?";
|
||||
system($synccmd) == 0 or do {
|
||||
warn "CRITICAL ERROR: $synccmd failed: $?";
|
||||
if ($exitcode < 2) { $exitcode = 2; }
|
||||
return 0;
|
||||
};
|
||||
|
||||
# now do an -I to the new sync snapshot, assuming there were any snapshots
|
||||
# other than the new sync snapshot to begin with, of course - and that we
|
||||
|
|
@ -302,6 +351,7 @@ sub syncdataset {
|
|||
# make sure target is (still) not currently in receive.
|
||||
if (iszfsbusy($targethost,$targetfs,$targetisroot)) {
|
||||
warn "Cannot sync now: $targetfs is already target of a zfs receive process.\n";
|
||||
if ($exitcode < 1) { $exitcode = 1; }
|
||||
return 0;
|
||||
}
|
||||
|
||||
|
|
@ -309,9 +359,12 @@ sub syncdataset {
|
|||
if ($debug) { print "DEBUG: $synccmd\n"; }
|
||||
|
||||
if ($oldestsnap ne $newsyncsnap) {
|
||||
system($synccmd) == 0
|
||||
or warn "CRITICAL ERROR: $synccmd failed: $?";
|
||||
my $ret = system($synccmd);
|
||||
if ($ret != 0) {
|
||||
warn "CRITICAL ERROR: $synccmd failed: $?";
|
||||
if ($exitcode < 1) { $exitcode = 1; }
|
||||
return 0;
|
||||
}
|
||||
} else {
|
||||
if (!$quiet) { print "INFO: no incremental sync needed; $oldestsnap is already the newest available snapshot.\n"; }
|
||||
}
|
||||
|
|
@ -335,8 +388,11 @@ sub syncdataset {
|
|||
|
||||
if (!$quiet) { print "Resuming interrupted zfs send/receive from $sourcefs to $targetfs (~ $disp_pvsize remaining):\n"; }
|
||||
if ($debug) { print "DEBUG: $synccmd\n"; }
|
||||
system("$synccmd") == 0
|
||||
or die "CRITICAL ERROR: $synccmd failed: $?";
|
||||
system("$synccmd") == 0 or do {
|
||||
warn "CRITICAL ERROR: $synccmd failed: $?";
|
||||
if ($exitcode < 2) { $exitcode = 2; }
|
||||
return 0;
|
||||
};
|
||||
|
||||
# a resumed transfer will only be done to the next snapshot,
|
||||
# so do an normal sync cycle
|
||||
|
|
@ -364,6 +420,7 @@ sub syncdataset {
|
|||
# make sure target is (still) not currently in receive.
|
||||
if (iszfsbusy($targethost,$targetfs,$targetisroot)) {
|
||||
warn "Cannot sync now: $targetfs is already target of a zfs receive process.\n";
|
||||
if ($exitcode < 1) { $exitcode = 1; }
|
||||
return 0;
|
||||
}
|
||||
|
||||
|
|
@ -393,16 +450,20 @@ sub syncdataset {
|
|||
if ($debug) { print "DEBUG: $synccmd\n"; }
|
||||
|
||||
my $output = `$synccmd 2>&1`;
|
||||
my $exitcode = $?;
|
||||
my $ret = $?;
|
||||
|
||||
if ($exitcode != 0) {
|
||||
if ($ret != 0) {
|
||||
if (!$resume && $output =~ /\Qcontains partially-complete state\E/) {
|
||||
if (!$quiet) { print "Resetting partially receive state\n"; }
|
||||
resetreceivestate($targethost,$targetfs,$targetisroot);
|
||||
system($synccmd) == 0
|
||||
or die "CRITICAL ERROR: $synccmd failed: $?";
|
||||
system("$synccmd") == 0 or do {
|
||||
warn "CRITICAL ERROR: $synccmd failed: $?";
|
||||
if ($exitcode < 2) { $exitcode = 2; }
|
||||
return 0;
|
||||
};
|
||||
} else {
|
||||
die "CRITICAL ERROR: $synccmd failed: $exitcode";
|
||||
warn "CRITICAL ERROR: $synccmd failed: $ret";
|
||||
if ($exitcode < 2) { $exitcode = 2; }
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -574,7 +635,7 @@ sub checkcommands {
|
|||
if ($debug) { print "DEBUG: checking availability of $mbuffercmd on source...\n"; }
|
||||
$avail{'sourcembuffer'} = `$sourcessh $lscmd $mbuffercmd 2>/dev/null`;
|
||||
if ($avail{'sourcembuffer'} eq '') {
|
||||
print "WARN: $mbuffercmd not available on source $s - sync will continue without source buffering.\n";
|
||||
if (!$quiet) { print "WARN: $mbuffercmd not available on source $s - sync will continue without source buffering.\n"; }
|
||||
$avail{'sourcembuffer'} = 0;
|
||||
} else {
|
||||
$avail{'sourcembuffer'} = 1;
|
||||
|
|
@ -583,7 +644,7 @@ sub checkcommands {
|
|||
if ($debug) { print "DEBUG: checking availability of $mbuffercmd on target...\n"; }
|
||||
$avail{'targetmbuffer'} = `$targetssh $lscmd $mbuffercmd 2>/dev/null`;
|
||||
if ($avail{'targetmbuffer'} eq '') {
|
||||
print "WARN: $mbuffercmd not available on target $t - sync will continue without target buffering.\n";
|
||||
if (!$quiet) { print "WARN: $mbuffercmd not available on target $t - sync will continue without target buffering.\n"; }
|
||||
$avail{'targetmbuffer'} = 0;
|
||||
} else {
|
||||
$avail{'targetmbuffer'} = 1;
|
||||
|
|
@ -595,14 +656,14 @@ sub checkcommands {
|
|||
$avail{'localmbuffer'} = `$lscmd $mbuffercmd 2>/dev/null`;
|
||||
if ($avail{'localmbuffer'} eq '') {
|
||||
$avail{'localmbuffer'} = 0;
|
||||
print "WARN: $mbuffercmd not available on local machine - sync will continue without local buffering.\n";
|
||||
if (!$quiet) { print "WARN: $mbuffercmd not available on local machine - sync will continue without local buffering.\n"; }
|
||||
}
|
||||
}
|
||||
|
||||
if ($debug) { print "DEBUG: checking availability of $pvcmd on local machine...\n"; }
|
||||
$avail{'localpv'} = `$lscmd $pvcmd 2>/dev/null`;
|
||||
if ($avail{'localpv'} eq '') {
|
||||
print "WARN: $pvcmd not available on local machine - sync will continue without progress bar.\n";
|
||||
if (!$quiet) { print "WARN: $pvcmd not available on local machine - sync will continue without progress bar.\n"; }
|
||||
$avail{'localpv'} = 0;
|
||||
} else {
|
||||
$avail{'localpv'} = 1;
|
||||
|
|
@ -610,7 +671,7 @@ sub checkcommands {
|
|||
|
||||
# check for ZFS resume feature support
|
||||
if ($resume) {
|
||||
my $resumechkcmd = "$zfscmd get receive_resume_token -d 0";
|
||||
my $resumechkcmd = "$zfscmd get -d 0 receive_resume_token";
|
||||
|
||||
if ($debug) { print "DEBUG: checking availability of zfs resume feature on source...\n"; }
|
||||
$avail{'sourceresume'} = system("$sourcessh $resumechkcmd >/dev/null 2>&1");
|
||||
|
|
@ -730,7 +791,7 @@ sub getoldestsnapshot {
|
|||
# must not have had any snapshots on source - luckily, we already made one, amirite?
|
||||
if (defined ($args{'no-sync-snap'}) ) {
|
||||
# well, actually we set --no-sync-snap, so no we *didn't* already make one. Whoops.
|
||||
die "CRIT: --no-sync-snap is set, and getoldestsnapshot() could not find any snapshots on source!\n";
|
||||
warn "CRIT: --no-sync-snap is set, and getoldestsnapshot() could not find any snapshots on source!\n";
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
|
|
@ -739,7 +800,7 @@ sub getnewestsnapshot {
|
|||
my $snaps = shift;
|
||||
foreach my $snap ( sort { $snaps{'source'}{$b}{'creation'}<=>$snaps{'source'}{$a}{'creation'} } keys %{ $snaps{'source'} }) {
|
||||
# return on first snap found - it's the newest
|
||||
print "NEWEST SNAPSHOT: $snap\n";
|
||||
if (!$quiet) { print "NEWEST SNAPSHOT: $snap\n"; }
|
||||
return $snap;
|
||||
}
|
||||
# must not have had any snapshots on source - looks like we'd better create one!
|
||||
|
|
@ -752,6 +813,7 @@ sub getnewestsnapshot {
|
|||
# we also probably need an argument to mute this WARN, for people who deliberately exclude
|
||||
# datasets from recursive replication this way.
|
||||
warn "WARN: --no-sync-snap is set, and getnewestsnapshot() could not find any snapshots on source for current dataset. Continuing.\n";
|
||||
if ($exitcode < 2) { $exitcode = 2; }
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
|
|
@ -769,7 +831,7 @@ sub buildsynccmd {
|
|||
$synccmd = "$sendcmd |";
|
||||
# avoid confusion - accept either source-bwlimit or target-bwlimit as the bandwidth limiting option here
|
||||
my $bwlimit = '';
|
||||
if (length $args{'bwlimit'}) {
|
||||
if (length $args{'source-bwlimit'}) {
|
||||
$bwlimit = $args{'source-bwlimit'};
|
||||
} elsif (length $args{'target-bwlimit'}) {
|
||||
$bwlimit = $args{'target-bwlimit'};
|
||||
|
|
@ -850,7 +912,7 @@ sub pruneoldsyncsnaps {
|
|||
|
||||
# only prune snaps beginning with syncoid and our own hostname
|
||||
foreach my $snap(@snaps) {
|
||||
if ($snap =~ /^syncoid_\Q$hostid\E/) {
|
||||
if ($snap =~ /^syncoid_\Q$identifier$hostid\E/) {
|
||||
# no matter what, we categorically refuse to
|
||||
# prune the new sync snap we created for this run
|
||||
if ($snap ne $newsyncsnap) {
|
||||
|
|
@ -875,7 +937,7 @@ sub pruneoldsyncsnaps {
|
|||
$prunecmd = escapeshellparam($prunecmd);
|
||||
}
|
||||
system("$rhost $prunecmd") == 0
|
||||
or warn "CRITICAL ERROR: $rhost $prunecmd failed: $?";
|
||||
or warn "WARNING: $rhost $prunecmd failed: $?";
|
||||
$prunecmd = '';
|
||||
$counter = 0;
|
||||
}
|
||||
|
|
@ -906,6 +968,7 @@ sub getmatchingsnapshot {
|
|||
}
|
||||
|
||||
# if we got this far, we failed to find a matching snapshot.
|
||||
if ($exitcode < 2) { $exitcode = 2; }
|
||||
|
||||
print "\n";
|
||||
print "CRITICAL ERROR: Target $targetfs exists but has no snapshots matching with $sourcefs!\n";
|
||||
|
|
@ -936,10 +999,14 @@ sub newsyncsnap {
|
|||
if ($isroot) { $mysudocmd = ''; } else { $mysudocmd = $sudocmd; }
|
||||
my $hostid = hostname();
|
||||
my %date = getdate();
|
||||
my $snapname = "syncoid\_$hostid\_$date{'stamp'}";
|
||||
my $snapname = "syncoid\_$identifier$hostid\_$date{'stamp'}";
|
||||
my $snapcmd = "$rhost $mysudocmd $zfscmd snapshot $fsescaped\@$snapname\n";
|
||||
system($snapcmd) == 0
|
||||
or die "CRITICAL ERROR: $snapcmd failed: $?";
|
||||
system($snapcmd) == 0 or do {
|
||||
warn "CRITICAL ERROR: $snapcmd failed: $?";
|
||||
if ($exitcode < 2) { $exitcode = 2; }
|
||||
return 0;
|
||||
};
|
||||
|
||||
return $snapname;
|
||||
}
|
||||
|
||||
|
|
@ -1194,7 +1261,9 @@ syncoid - ZFS snapshot replication tool
|
|||
Options:
|
||||
|
||||
--compress=FORMAT Compresses data during transfer. Currently accepted options are gzip, pigz-fast, pigz-slow, lzo (default) & none
|
||||
--identifier=EXTRA Extra identifier which is included in the snapshot name. Can be used for replicating to multiple targets.
|
||||
--recursive|r Also transfers child datasets
|
||||
--skip-parent Skips syncing of the parent dataset. Does nothing without '--recursive' option.
|
||||
--source-bwlimit=<limit k|m|g|t> Bandwidth limit on the source transfer
|
||||
--target-bwlimit=<limit k|m|g|t> Bandwidth limit on the target transfer
|
||||
--no-stream Replicates using newest snapshot instead of intermediates
|
||||
|
|
|
|||
Loading…
Reference in New Issue