This prevents unnecessary message renumbering and I/O.

Without this change, there is a small window for long-running
WWW streaming requests to miss a message that was unref-ed
before reindexing.  If we expose an "All Mail" mailbox via
IMAP/JMAP, this will save client traffic.
---
 lib/PublicInbox/ExtSearchIdx.pm | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/lib/PublicInbox/ExtSearchIdx.pm b/lib/PublicInbox/ExtSearchIdx.pm
index c2ab0447e176..40489eab4c66 100644
--- a/lib/PublicInbox/ExtSearchIdx.pm
+++ b/lib/PublicInbox/ExtSearchIdx.pm
@@ -193,6 +193,7 @@ sub do_xpost ($$) {
                $idx->ipc_do('add_eidx_info', $docid, $eidx_key, $eml);
                apply_boost($req, $smsg) if $req->{boost_in_use};
        } else { # 'd' no {xnum}
+               $self->git->async_wait_all;
                $oid = pack('H*', $oid);
                _unref_doc($req, $docid, $xibx, undef, $oid, $eml);
        }
@@ -261,6 +262,7 @@ sub _blob_missing ($$) { # called when a known 
$smsg->{blob} is gone
        # xnum and ibx are unknown, we only call this when an entry from
        # /ei*/over.sqlite3 is bad, not on entries from xap*/over.sqlite3
        my $oidbin = pack('H*', $smsg->{blob});
+       $req->{self}->git->async_wait_all;
        _unref_doc($req, $smsg, undef, undef, $oidbin);
 }
 
@@ -552,6 +554,7 @@ sub _reindex_finalize ($$$) {
        }
        return if $nr == 1; # likely, all good
 
+       $self->git->async_wait_all;
        warn "W: #$docid split into $nr due to deduplication change\n";
        my @todo;
        for my $ary (values %$by_chash) {
@@ -896,6 +899,7 @@ ibx_id = ? AND xnum >= ? AND xnum <= ?
                }
                return if $sync->{quit};
                next unless scalar keys %x3m;
+               $self->git->async_wait_all; # wait for reindex_unseen
 
                # eliminate stale/mismatched entries
                my %mismatch = map { $_->{num} => $_->{blob} } @$msgs;
--
unsubscribe: one-click, see List-Unsubscribe header
archive: https://public-inbox.org/meta/

Reply via email to