[pve-devel] [PATCH v5 manager 1/2] vzdump: move remaining guest include logic to single method
Aaron Lauterer
a.lauterer at proxmox.com
Mon Jun 8 13:55:27 CEST 2020
On 6/5/20 8:39 PM, Thomas Lamprecht wrote:
> On 5/6/20 11:57 AM, Aaron Lauterer wrote:
>>
[...]
>> diff --git a/PVE/API2/VZDump.pm b/PVE/API2/VZDump.pm
>> index 68a3de89..77dac1b1 100644
>> --- a/PVE/API2/VZDump.pm
>> +++ b/PVE/API2/VZDump.pm
>> @@ -69,19 +69,26 @@ __PACKAGE__->register_method ({
>> return 'OK' if $param->{node} && $param->{node} ne $nodename;
>>
>> my $cmdline = PVE::VZDump::Common::command_line($param);
>> - my ($vmids, $skiplist) = PVE::VZDump->get_included_guests($param);
>> +
>> + my $vmids_per_node = PVE::VZDump->get_included_guests($param);
>> +
>> + my $vmids = $vmids_per_node->{$nodename} // [];
>
> maybe it could make belows looping easier if you delete the local ones out?
> So, for above:
> my $local_vmids = delete $vmids_per_node->{$nodename} // [];
>
>> +
>> + my $skiplist = [];
>> +
>> + for my $current_node (keys %{$vmids_per_node}) {
>> + push @{$skiplist}, @{$vmids_per_node->{$current_node}}
>> + if $current_node ne $nodename;
>
> with the delete above you could drop the post-if here actually you should be
> able to just do:
>
> my $skiplist = [ map { @$_ } values $vmids_per_node->%* ];
Thanks for that idea :)
>
>
> (FYI: $vmids_per_node->%* is just alternate writing of %$vmids_per_node IIRC,
> Wolfgang prefers it)
>
>> + }
>>
>> if($param->{stop}){
>> PVE::VZDump::stop_running_backups();
>> return 'OK' if !scalar(@{$vmids});
>> }
>>
>> - # silent exit if specified VMs run on other nodes
>> + # silent exit if specified guests run on other nodes
>> return "OK" if !scalar(@{$vmids});
>>
>> - my @exclude = PVE::Tools::split_list(extract_param($param, 'exclude'));
>> - $param->{exclude} = PVE::VZDump::check_vmids(@exclude);
>> -
>> # exclude-path list need to be 0 separated
>> if (defined($param->{'exclude-path'})) {
>> my @expaths = split(/\0/, $param->{'exclude-path'} || '');
>> @@ -106,6 +113,7 @@ __PACKAGE__->register_method ({
>> die "interrupted by signal\n";
>> };
>>
>> + $param->{vmids} = $vmids;
>> my $vzdump = PVE::VZDump->new($cmdline, $param, $skiplist);
>>
>> eval {
>> diff --git a/PVE/VZDump.pm b/PVE/VZDump.pm
>> index 68c08f47..cd278f4d 100644
>> --- a/PVE/VZDump.pm
>> +++ b/PVE/VZDump.pm
>> @@ -1039,29 +1039,18 @@ sub exec_backup {
>> if scalar(@{$self->{skiplist}});
>>
>> my $tasklist = [];
>> + my $vzdump_plugins = {};
>> + foreach my $plugin (@{$self->{plugins}}) {
>
> maybe add:
> next if exists $vzdump_plugins->{$type};
>
> as then we'd just overwrite the same type again.
yeah sounds good.
>
>> + my $type = $plugin->type();
>> + $vzdump_plugins->{$type} = $plugin;
>> + }
>>
>> - if ($opts->{all}) {
>> - foreach my $plugin (@{$self->{plugins}}) {
>> - my $vmlist = $plugin->vmlist();
>> - foreach my $vmid (sort @$vmlist) {
>> - next if grep { $_ eq $vmid } @{$opts->{exclude}};
>> - next if !$rpcenv->check($authuser, "/vms/$vmid", [ 'VM.Backup' ], 1);
>> - push @$tasklist, { vmid => $vmid, state => 'todo', plugin => $plugin, mode => $opts->{mode} };
>> - }
>> - }
>> - } else {
>> - foreach my $vmid (sort @{$opts->{vmids}}) {
>> - my $plugin;
>> - foreach my $pg (@{$self->{plugins}}) {
>> - my $vmlist = $pg->vmlist();
>> - if (grep { $_ eq $vmid } @$vmlist) {
>> - $plugin = $pg;
>> - last;
>> - }
>> - }
>> - $rpcenv->check($authuser, "/vms/$vmid", [ 'VM.Backup' ]);
>> - push @$tasklist, { vmid => $vmid, state => 'todo', plugin => $plugin, mode => $opts->{mode} };
>> - }
>> + my $vmlist = PVE::Cluster::get_vmlist();
>> + foreach my $vmid (sort @{$opts->{vmids}}) {
>> + my $guest_type = $vmlist->{ids}->{$vmid}->{type};
>> + my $plugin = $vzdump_plugins->{$guest_type};
>> + next if !$rpcenv->check($authuser, "/vms/$vmid", [ 'VM.Backup' ], $opts->{all});
>> + push @$tasklist, { vmid => $vmid, state => 'todo', plugin => $plugin, mode => $opts->{mode} };
>
> I now it blows up lines a bit but I prefer having multiple hash elements on
> separate ones most of the time.
>
> push @$tasklist, {
> vmid => $vmid,
> ...,
> };
>
> Easier to read and change.
hehe, I just read our coding guidelines a few days ago and that
particular thing was one of my personal take aways ;)
>
>> }
>>
>> # Use in-memory files for the outer hook logs to pass them to sendmail.
>> @@ -1169,34 +1158,39 @@ sub get_included_guests {
>>
>> my $nodename = PVE::INotify::nodename();
>> my $vmids = [];
>> + my $vmids_per_node = {};
>> +
>> + my $vmlist = PVE::Cluster::get_vmlist();
>>
>> - # convert string lists to arrays
>> if ($job->{pool}) {
>> $vmids = PVE::API2Tools::get_resource_pool_guest_members($job->{pool});
>> + } elsif ($job->{vmid}) {
>> + $vmids = [ PVE::Tools::split_list($job->{vmid}) ];
>> } else {
>> - $vmids = [ PVE::Tools::split_list(extract_param($job, 'vmid')) ];
>> + # all or exclude
>> + my @exclude = PVE::Tools::split_list($job->{exclude});
>> + @exclude = @{PVE::VZDump::check_vmids(@exclude)};
>> + my $excludehash = { map { $_ => 1 } @exclude };
>> +
>> + for my $id (keys %{ $vmlist->{ids} }) {
>
> you could sort it here already, avoiding the sort after the loop.
>
> Whole thing could also be written as:
>
> $vmids = [ grep { !$excludehash->{$_} } sort keys $vmlist->{ids}->%* ];
>
> but no hard feeling here.
I agree on the early sorting, but I am not sure about that oneliner.
It's not as easy to comprehend what's going on for someone who isn't
used to this kind of syntax already.
>
>> + next if $excludehash->{$id};
>> + push @$vmids, $id;
>> + }
>> }
>> + $vmids = [ sort {$a <=> $b} @$vmids];
>>
>> - my $skiplist = [];
>> - if (!$job->{all}) {
>> - if (!$job->{node} || $job->{node} eq $nodename) {
>> - my $vmlist = PVE::Cluster::get_vmlist();
>> - my $localvmids = [];
>> - foreach my $vmid (@{$vmids}) {
>> - my $d = $vmlist->{ids}->{$vmid};
>> - if ($d && ($d->{node} ne $nodename)) {
>> - push @{$skiplist}, $vmid;
>> - } else {
>> - push @{$localvmids}, $vmid;
>> - }
>> - }
>> - $vmids = $localvmids;
>> - }
>> + $vmids = PVE::VZDump::check_vmids(@$vmids);
>> +
>> + foreach my $vmid (@$vmids) {
>> + my $vmid_data = $vmlist->{ids}->{$vmid};
>> + my $node = $vmid_data->{node};
>> +
>> + next if (defined $job->{node} && $job->{node} ne $node);>
>> - $job->{vmids} = PVE::VZDump::check_vmids(@{$vmids})
>> + push @{$vmids_per_node->{$node}}, $vmid;
>> }
>>
>> - return ($vmids, $skiplist);
>> + return $vmids_per_node;
>> }
>>
>> 1;
>>
>
More information about the pve-devel
mailing list