MIME-Version: 1.0
From: Ashish Mukherjee <ashish.mukherjee@gmail.com>
Date: Thu, 25 Sep 2025 17:06:33 +0530
Message-ID: <CACgMzfxGyCQHhE3BOrn0c8GpAd0ZMQGc80_C-qYUWZArEKqbZg@mail.gmail.com>
Subject: Enquiry about Percona TDE performance issues
To: pgsql-general@postgresql.org
Content-Type: multipart/alternative; boundary="0000000000008ae815063f9e95f5"
Archived-At: <https://www.postgresql.org/message-id/CACgMzfxGyCQHhE3BOrn0c8GpAd0ZMQGc80_C-qYUWZArEKqbZg%40mail.gmail.com>
Precedence: bulk

--0000000000008ae815063f9e95f5
Content-Type: text/plain; charset="UTF-8"

Hello,

I recently upgraded from pgsql 12 to pgsql 17. It's a Percona TDE enabled
database. However, after the upgrade the query planning time has shot up by
10x when I check explain analyze output.

The difference in planning time is visible here. The first case is pgsql 17
with TDE. The second case pgsql 12 with TDE.

s14=> explain analyze select invoicenum from ltlinvoice where
pseq=1400000224216382;
                                                          QUERY PLAN
------------------------------------------------------------------------------------------------------------------------------
 Index Scan using ltlinvoice_pseq on ltlinvoice  (cost=0.43..10.10 rows=9
width=12) (actual time=0.040..0.040 rows=0 loops=1)
   Index Cond: (pseq = '1400000224216382'::bigint)
 Planning Time: 158.199 ms
 Execution Time: 0.051 ms

s14=> explain analyze select invoicenum from ltlinvoice where
pseq=1400000224216382;
                                                          QUERY PLAN
------------------------------------------------------------------------------------------------------------------------------
 Index Scan using ltlinvoice_pseq on ltlinvoice  (cost=0.43..10.10 rows=9
width=12) (actual time=0.023..0.023 rows=0 loops=1)
   Index Cond: (pseq = '1400000224216382'::bigint)
 Planning Time: 0.195 ms
 Execution Time: 0.046 ms

Any pointers about this would be helpful.

Regards,
Ashish

--0000000000008ae815063f9e95f5
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr">Hello,<div><br></div><div>I recently upgraded from pgsql 1=
2 to pgsql 17. It&#39;s a Percona TDE enabled database. However, after the =
upgrade the query planning time has shot up by 10x when I check explain ana=
lyze output.=C2=A0</div><div><br></div><div>The difference in planning time=
 is visible here. The first case is pgsql 17 with TDE. The second case pgsq=
l 12 with TDE.</div><div><br></div><div>s14=3D&gt; explain analyze select i=
nvoicenum from ltlinvoice where pseq=3D1400000224216382;<br>=C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 QUERY PLAN<br>-------------------=
---------------------------------------------------------------------------=
--------------------------------<br>=C2=A0Index Scan using ltlinvoice_pseq =
on ltlinvoice =C2=A0(cost=3D0.43..10.10 rows=3D9 width=3D12) (actual time=
=3D0.040..0.040 rows=3D0 loops=3D1)<br>=C2=A0 =C2=A0Index Cond: (pseq =3D &=
#39;1400000224216382&#39;::bigint)<br>=C2=A0Planning Time: 158.199 ms<br>=
=C2=A0Execution Time: 0.051 ms<br><br>s14=3D&gt; explain analyze select inv=
oicenum from ltlinvoice where pseq=3D1400000224216382;<br>=C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=
=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 QUERY PLAN<br>----------------------=
---------------------------------------------------------------------------=
-----------------------------<br>=C2=A0Index Scan using ltlinvoice_pseq on =
ltlinvoice =C2=A0(cost=3D0.43..10.10 rows=3D9 width=3D12) (actual time=3D0.=
023..0.023 rows=3D0 loops=3D1)<br>=C2=A0 =C2=A0Index Cond: (pseq =3D &#39;1=
400000224216382&#39;::bigint)<br>=C2=A0Planning Time: 0.195 ms<br>=C2=A0Exe=
cution Time: 0.046 ms</div><div><br></div><div>Any pointers about this woul=
d be helpful.</div><div><br></div><div>Regards,</div><div>Ashish</div></div=
>

--0000000000008ae815063f9e95f5--