MIME-Version: 1.0
References: <CAGqGHCNs_OWZjhNghiaENCLdRF3g7zc4cCegRcYwB1RFLKVf8g@mail.gmail.com>
 <93e769c5-e5d8-411d-b559-dc41534c1bc6@aklaver.com> <CAGqGHCNazFXwCo=EB2UAx5kZ771fKpYDzD41v9RPVUtb9Z3JSg@mail.gmail.com>
In-Reply-To: <CAGqGHCNazFXwCo=EB2UAx5kZ771fKpYDzD41v9RPVUtb9Z3JSg@mail.gmail.com>
From: =?UTF-8?Q?Torsten_F=C3=B6rtsch?= <tfoertsch123@gmail.com>
Date: Fri, 24 Jan 2025 12:07:18 +0100
Message-ID: <CAKkG4_m_O_=o8uBQu3ox-ZxF61mzHLSB+Q6GkvHfyAQyPRZmkw@mail.gmail.com>
Subject: Re: Is postgresql's json strong consistency or eventually consistency?
To: anlex N <an0291170@gmail.com>
Cc: Adrian Klaver <adrian.klaver@aklaver.com>, pgsql-general@lists.postgresql.org
Content-Type: multipart/alternative; boundary="0000000000008a2408062c71bb72"
Archived-At: <https://www.postgresql.org/message-id/CAKkG4_m_O_%3Do8uBQu3ox-ZxF61mzHLSB%2BQ6GkvHfyAQyPRZmkw%40mail.gmail.com>
Precedence: bulk

--0000000000008a2408062c71bb72
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Fri, Jan 24, 2025 at 4:48=E2=80=AFAM anlex N <an0291170@gmail.com> wrote=
:

> Hello Adrian, Laurenz. Have you tried postgresql's json in your everyday
> life? How fast is it? how is it than mongodb?
>

My honest opinion, JSON(B) in PG is great as data transfer format but not
so much for storage. Here is an example. Up to last year I had a table in
my DB with rows consisting of a bunch of numbers in many columns. Over
maybe 15 years of business that table had grown to a few TB in size. Then
the development team decided they needed more flexibility and started to
fill a new table with the same information in JSON. After only 1 year of
business that table is now in the same size range.

Another example, the developers came up with a new table to store the
result of the JSON response they got from a 3rd party application. That
worked for a while and the table behaved normally. Suddenly it started
growing by a few GB a day. Turns out the 3rd party had decided to include
an MB-sized picture in the JSON response. Our team then faithfully stored
all of that crap unfiltered in the DB.

That is not an answer to your original question, of course. How fast it is
depends very much on the use case. If you are talking about access methods
alone, the ability of indexing JSON and such, then PG is on par with mongo
if not better. But nobody but you can give you a definite answer relating
to your situation.

--0000000000008a2408062c71bb72
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div dir=3D"ltr"><br></div><div class=3D"gmail_quote gmail=
_quote_container"><div dir=3D"ltr" class=3D"gmail_attr">On Fri, Jan 24, 202=
5 at 4:48=E2=80=AFAM anlex N &lt;<a href=3D"mailto:an0291170@gmail.com">an0=
291170@gmail.com</a>&gt; wrote:<br></div><blockquote class=3D"gmail_quote" =
style=3D"margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);pa=
dding-left:1ex"><div dir=3D"ltr"><div style=3D"font-family:monospace,monosp=
ace">Hello=C2=A0Adrian,=C2=A0Laurenz. Have you tried postgresql&#39;s json =
in your everyday life? How fast is it? how is it than mongodb?</div></div><=
/blockquote><div><br></div><div>My honest opinion, JSON(B) in PG is great a=
s data transfer format but not so much for storage. Here is an example. Up =
to last year I had a table in my DB with rows consisting of a bunch=C2=A0of=
 numbers in many columns. Over maybe 15 years of business that table had gr=
own to a few TB in size. Then the development team decided they needed more=
 flexibility and started to fill a new table with the same information in J=
SON. After only 1 year of business that table is now in the same size range=
.</div><div><br></div><div>Another example, the developers came up with a n=
ew table to store the result of the JSON response they got from a 3rd party=
 application. That worked for a while and the table behaved normally. Sudde=
nly it started growing by a few GB a day. Turns out the 3rd party had decid=
ed to include an MB-sized picture in the JSON response. Our team then faith=
fully stored all of that crap unfiltered in the DB.</div><div><br></div><di=
v>That is not an answer to your original question, of course. How fast it i=
s depends very much on the use case. If you are talking about access method=
s alone, the ability of indexing JSON and such, then PG is on par with mong=
o if not better. But nobody=C2=A0but you=C2=A0can give you a definite answe=
r relating to your situation.</div></div></div>

--0000000000008a2408062c71bb72--