From 45095845fd4b88689f74c3eecbb3021151002d30 Mon Sep 17 00:00:00 2001
From: "Peter J. Holzer" <hjp@hjp.at>
Date: Wed, 29 Nov 2023 06:04:02 +0100
Subject: [PATCH] Limit instances to deduplicate by id

The instances table is partitioned by id. If we restrinct the search to
recent ids we only have to search a few partitions which should be
faster.
---
 lib/Simba/CA.pm | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/lib/Simba/CA.pm b/lib/Simba/CA.pm
index 956cefc..6f04022 100644
--- a/lib/Simba/CA.pm
+++ b/lib/Simba/CA.pm
@@ -727,6 +727,9 @@ sub linkdup {
             push @{ $self->{sessions_on_same_device} }, $r->{id}; # We only use the id
         }
 
+        my $self->{min_instance_id} = $self->{dbh}->selectrow_array("select min(id) from instances where session in " . join(", ", map("?", @{ $self->{sessions_on_same_device} })) . ")");
+        $self->log(3, "min_instance_id set to $self->{min_instance_id}");
+
     }
     my $tdb0 = gettimeofday();
     my $tdb1;
@@ -735,8 +738,9 @@ sub linkdup {
                                        and file_owner=? and file_group=? and file_acl=?
                                        and file_unix_bits=?
                                        and checksum=? and online=1
-                                       and instances.file=files.id
                                        and versions2.id=instances.version
+                                       and instances.id >= ?
+                                       and instances.file=files.id
                                        and instances.session=sessions.id
                                        and sessions.id in (" . join(", ", map("?", @{ $self->{sessions_on_same_device} })) . ")" .
                                     " order by instances.session desc
@@ -746,6 +750,7 @@ sub linkdup {
              $f->{o}, $f->{g}, $f->{acl},
              join(',', map {$f->{$_} ? ($_) : ()} qw(setuid setgid sticky)),
              $f->{checksum},
+             $self->{min_instance_id},
              @{ $self->{sessions_on_same_device} },
           );
     my $st = stat($backup_filename);