Postgres Professional
   

Re: TCP network cost

From: PFC
Subject: Re: TCP network cost
Date: ,
Msg-id: op.upln5emacigqcu@soyouz
(view: Whole thread, Raw)
In response to: Re: TCP network cost  ("Ross J. Reedstrom")
Responses: Re: TCP network cost  ("Ross J. Reedstrom")
List: pgsql-performance

Tree view

TCP network cost  ("Ross J. Reedstrom", )
 Re: TCP network cost  (Rusty Conover, )
  Re: TCP network cost  (, )
  Re: TCP network cost  ("Ross J. Reedstrom", )
   Re: TCP network cost  (Rusty Conover, )
    Re: TCP network cost  ("Ross J. Reedstrom", )
     Re: TCP network cost  ("Ross J. Reedstrom", )
      Re: TCP network cost  (Aaron Turner, )
     Re: TCP network cost  (PFC, )
      Re: TCP network cost  ("Ross J. Reedstrom", )
   Re: TCP network cost  (Gregory Stark, )
    Re: TCP network cost  ("Ross J. Reedstrom", )
     Re: TCP network cost  (Tom Lane, )
      Re: TCP network cost  ("Ross J. Reedstrom", )
       Re: TCP network cost  (Linos, )
        Re: TCP network cost  (Tom Lane, )
         Re: TCP network cost  (Linos, )
          Re: TCP network cost  (Tom Lane, )
           Re: TCP network cost  (Linos, )
            Re: TCP network cost  (Tom Lane, )
             Re: TCP network cost  (Magnus Hagander, )
              Re: TCP network cost  (Linos, )
 Re: TCP network cost  (Aaron Turner, ) 
> python w/ psycopg (or psycopg2), which wraps libpq. Same results w/
> either version.

    I've seen psycopg2 saturate a 100 Mbps ethernet connection (direct
connection with crossover cable) between postgres server and client during
a benchmark... I had to change the benchmark to not retrieve a large TEXT
column to remove this bottleneck... this was last year so versions are
probably different, but I don't think this matters a lot...

> Note the 'bare' transfer times added above. Nothing to write home about
> (~3Mb/sec) but another order of magnitude faster than the postgresql
> transfer.

    You should test with sending a large (>100 MB) amount of data through
Netcat. This should give you your maximum wire speed. Use /dev/null as the
test file, and use "pv" (pipe viewer) to measure throughput :

box 1 : pv < /dev/zero | nc -lp 12345
box 2 : nc (ip) 12345 >/dev/null

    On gigabit lan you should get 100 MB/s, on 100BaseT about 10 MB/s. If you
dont get that, there is a problem somewhere (bad cable, bad NIC, slow
switch/router, etc). Monitor CPU during this test (vmstat). Usage should
be low.

pgsql-performance by date:

Previous:
From: PFC
Date:
Subject: Re: TCP network cost
Next:
From: Grzegorz Jaśkiewicz
Date:
Subject: not in(subselect) in 8.4