As Greg Smith rightly said, i faced a problem of missing connections between the runs. I even ran the cron every less than a second, but, still that would become too many runs per second and later i need to take the burden of calculating every thing from the log.
I did not really calculate the IO load while the logging is on. I would switch on "log_connections" and "log_disconnections" to log the number of connections and duration of a connection.
If i notice high IO's and huge log generation, then i think Greg Spileburg has suggested a good idea of using tcpdump on a different server. I would use this utility and see how it works (never used it before). Greg Spileburg, please help me with any sources of documents you have to use "tcpdump".
Thanks again and sorry for replying late on this !
Regards,
Venkat
On Thu, Aug 25, 2011 at 6:02 AM, MirrorX <mirrorx@xxxxxxxxx> wrote:
lately i did sth similar in one of our servers, to keep track of active, idle
and idle in transaction connections so as to make some optimization in the
connection pooling and i didn't notice any serious io activity there (had
the cron job run every minute). so imho unless the server is seriously io
bound at the moment, you won't notice any difference
--
View this message in context: http://postgresql.1045698.n5.nabble.com/How-to-track-number-of-connections-and-hosts-to-Postgres-cluster-tp4729546p4732518.html
Sent from the PostgreSQL - performance mailing list archive at Nabble.com.
--
Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-performance
