No subject
Fri Jan 30 20:15:43 JST 2015
there are many external connections sharing fewer connections to backend.
Instead, Pgpool accepts external connections and then multiply them to the
backend! I dont get what is the advantage when you've got 'permanent'
external connections (unless the cached connections can be shared among
every child).
Thanks for your work.
On 16 March 2015 at 04:56, Tatsuo Ishii <ishii at postgresql.org> wrote:
> I suspect you run out num_init_children and there's no free pgpool-II
> process to accept new connection. To make sure that you could use ps
> command. For pgpool-II to accept new connections from clients, there
> must be process which looks like something like this:
>
> 17167 pts/2 S 0:00 pgpool: wait for connection request
>
> Best regards,
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese:http://www.sraoss.co.jp
>
> > Hi!
> >
> > I've been using 3.3.4 in a new deployment and from time to time pgpool
> > doesn't reply anything to new PSQL connections.
> > Either with PgAdmin or a JDBC app or PSQL command line, the initial
> > handshake stays frozen. However we can connect to the Psql backends
> > directly throught their local ports.
> >
> > I've got pgpool right now with *replication mode = on* (but only 1
> active
> > node) and *'connection_cache = on'* with default values for pooling.
> >
> > I cannot guess what's the SQL query that is hanging pgpool process. Is
> > pgpool waiting for some some reply from PSQL? If it is so, why does it
> > block all other connections?
> >
> > Machine resources are OK (5% cpu and 10% RAM used)
> >
> > After 20-30min (not sure) Pgpool accepts again new PSQL connections.
> >
> > I am looking right now at pgpool's log with debug level = 1, but it is a
> > bit hard to understand the errors kind: T, C, etc... and so many lines
> per
> > second.
> > I'll keep looking though.
> >
> >
> > Thanks.
>
--089e0160a708207a2c05118c81bb
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
<div dir=3D"ltr">OK, thanks.<div><br></div><div>That was it.</div><div><br>=
</div><div>After reading documentarion I didn't get properly the connec=
tion pool from pgpool. Now I know "num_init_children" is the numb=
er of accepted connections from Psql clients.</div><div><br></div><div>From=
what I understand, Pgpool's pool is not similar to other pools where t=
here are many external connections sharing fewer connections to backend.</d=
iv><div>Instead, Pgpool accepts external connections and then multiply them=
to the backend! I dont get what is the advantage when you've got '=
permanent' external connections (unless the cached connections can be s=
hared among every child).</div><div><br></div><div>Thanks for your work.</d=
iv><div><br></div><div><br></div><div><br></div><div><br></div><div><br></d=
iv><div><br></div><div><br></div><div><br></div><div><br></div><div><br></d=
iv></div><div class=3D"gmail_extra"><br><div class=3D"gmail_quote">On 16 Ma=
rch 2015 at 04:56, Tatsuo Ishii <span dir=3D"ltr"><<a href=3D"mailto:ish=
ii at postgresql.org" target=3D"_blank">ishii at postgresql.org</a>></span> wr=
ote:<br><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border=
-left:1px #ccc solid;padding-left:1ex">I suspect you run out num_init_child=
ren and there's no free pgpool-II<br>
process to accept new connection. To make sure that you could use ps<br>
command. For pgpool-II to accept new connections from clients, there<br>
must be process which looks like something like this:<br>
<br>
17167 pts/2=C2=A0 =C2=A0 S=C2=A0 =C2=A0 =C2=A0 0:00 pgpool: wait for connec=
tion request<br>
<br>
Best regards,<br>
--<br>
Tatsuo Ishii<br>
SRA OSS, Inc. Japan<br>
English: <a href=3D"http://www.sraoss.co.jp/index_en.php" target=3D"_blank"=
>http://www.sraoss.co.jp/index_en.php</a><br>
Japanese:<a href=3D"http://www.sraoss.co.jp" target=3D"_blank">http://www.s=
raoss.co.jp</a><br>
<span class=3D""><br>
> Hi!<br>
><br>
> I've been using 3.3.4 in a new deployment and from time to time pg=
pool<br>
> doesn't reply anything to new PSQL connections.<br>
> Either with PgAdmin or a JDBC app or PSQL command line, the initial<br=
>
> handshake stays frozen. However we can connect to the Psql backends<br=
>
> directly throught their local ports.<br>
><br>
</span>> I've got pgpool=C2=A0 right now with *replication mode =3D =
on* (but only 1 active<br>
> node) and *'connection_cache =3D on'* with default values for =
pooling.<br>
<div class=3D"HOEnZb"><div class=3D"h5">><br>
> I cannot guess what's the SQL query that is hanging pgpool process=
. Is<br>
> pgpool waiting for some some reply from PSQL? If it is so, why does it=
<br>
> block all other connections?<br>
><br>
> Machine resources are OK (5% cpu and 10% RAM used)<br>
><br>
> After 20-30min (not sure) Pgpool accepts again new PSQL connections.<b=
r>
><br>
> I am looking right now at pgpool's log with debug level =3D 1, but=
it is a<br>
> bit hard to understand the errors kind: T, C, etc... and so many lines=
per<br>
> second.<br>
>=C2=A0 I'll keep=C2=A0 looking though.<br>
><br>
><br>
> Thanks.<br>
</div></div></blockquote></div><br></div>
--089e0160a708207a2c05118c81bb--
More information about the pgpool-general
mailing list