Re: multiple postgres processes after establishing tcp connection

On 7/31/15 5:52 AM, Andrey Lizenko wrote:
> If only simple telnet connection started:
>
>     [postgres@ubuntu12-vm][20150731 06:30:24]:/db2$ telnet 127.0.0.1 5551
>     Trying 127.0.0.1...
>     Connected to 127.0.0.1.
>     Escape character is '^]'.
>
> new postgres process appeared as a child of the first one.
>
>     [postgres@ubuntu12-vm][20150731 06:38:42]:/db2/master$ ps -ef | grep
>     "bin/postgres" | grep -v grep
>     postgres  5152     1  0 06:28 pts/0    00:00:00
>     /opt/postgres/9.4.4/bin/postgres -D /db2/master
>     postgres  6358  5152  0 06:39 ?        00:00:00
>     /opt/postgres/9.4.4/bin/postgres -D /db2/master
>
>
>
> Is this an attempt to create new worker?
> No records (presumably) in pg_stat_activity, only "invalid length of
> startup packet" after closing telnet session.

To be expected. There's a bunch of startup stuff that happens before a
backend shows up in pg_stat_activity. See BackendStartup() for details.
In particular, BackendInitialize() is what collects the startup packet
from the port before calling BackendRun() which calls PostgresMain()
which is what finally registers the backend in the proc array.
--
Jim Nasby, Data Architect, Blue Treble Consulting, Austin TX
Data in Trouble? Get it in Treble! http://BlueTreble.com