Received: from malur.postgresql.org ([217.196.149.56]) by arkaria.postgresql.org with esmtps (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.98.2) (envelope-from ) id 1x93Bm-00000001iLu-3GP9 for pgsql-general@arkaria.postgresql.org; Tue, 22 Sep 2026 16:17:51 +0000 Received: from localhost ([127.0.0.1] helo=malur.postgresql.org) by malur.postgresql.org with esmtp (Exim 4.98.2) (envelope-from ) id 1x93Am-0000000HPJi-47F2 for pgsql-general@arkaria.postgresql.org; Tue, 22 Sep 2026 16:16:48 +0000 Received: from makus.postgresql.org ([2001:4800:3e1:1::229]) by malur.postgresql.org with esmtps (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.98.2) (envelope-from ) id 1x93Am-0000000HPJa-2lmc for pgsql-general@lists.postgresql.org; Tue, 22 Sep 2026 16:16:48 +0000 Received: from sss.pgh.pa.us ([68.162.161.243]) by makus.postgresql.org with esmtps (TLS1.3) tls TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384 (Exim 4.98.2) (envelope-from ) id 1x93Ak-00000000hIN-1r1B for pgsql-general@postgresql.org; Tue, 22 Sep 2026 16:16:47 +0000 Received: from sss1.sss.pgh.pa.us (localhost [127.0.0.1]) by sss.pgh.pa.us (8.18.1/8.18.1) with ESMTP id 68MGGjCc1186123; Tue, 22 Sep 2026 12:16:45 -0400 From: Tom Lane To: Anthony Mallgren cc: pgsql-general@postgresql.org Subject: Re: Skipping Inserts In-reply-to: References: Comments: In-reply-to Anthony Mallgren message dated "Mon, 21 Sep 2026 17:30:25 -0400" MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <1186121.1790093805.1@sss.pgh.pa.us> Content-Transfer-Encoding: quoted-printable Date: Tue, 22 Sep 2026 12:16:45 -0400 Message-ID: <1186122.1790093805@sss.pgh.pa.us> List-Id: List-Help: List-Subscribe: List-Post: List-Owner: List-Archive: Archived-At: Precedence: bulk Anthony Mallgren writes: > I'm getting a lot of messages like this: > [warn] skipped a 3-word n-gram: table tid from new index tuple (2524,87) > cannot find insert offset between offsets 2 and 16 of block 36 in index > "congress_103_3_word_word1_idx" I don't know about the first part of that, maybe it's from your application? But the "table tid ..." part matches this: /* * postingoff should never be set more than once per leaf page * binary search. That would mean that there are duplicate ta= ble * TIDs in the index, which is never okay. Check for that her= e. */ if (insertstate->postingoff !=3D 0) ereport(ERROR, (errcode(ERRCODE_INDEX_CORRUPTED), errmsg_internal("table tid from new index tuple (= %u,%u) cannot find insert offset between offsets %u and %u of block %u in = index \"%s\"", ItemPointerGetBlockNumber(key->sc= antid), ItemPointerGetOffsetNumber(key->s= cantid), low, stricthigh, BufferGetBlockNumber(insertstate-= >buf), RelationGetRelationName(rel)))); So this code thinks your index is corrupt. REINDEX should help. regards, tom lane