Re: [ADMIN] Strange behavior of child table.

2011-06-01 Thread Greg Smith
On 06/01/2011 02:07 AM, Jenish wrote: select * from Round_Action where action_id =5 limit 100 → execution time 80 sec select * from Round_Action_CH1 action_id =5 limit 100 → execution time 0.1 sec First off: each of the child tables needs to have the index created on them. That

[ADMIN] Strange behavior of child table.

2011-05-31 Thread Jenish
Hi All, I have created partition on table Round_Action , which has 3 inherited partition tables. When I am firing a simple select query with limit on parent table it is taking huge time to execute. But when I am firing this query directly on inherited table it is taking few milliseconds. EXP.

Re: [ADMIN] Strange behavior

2007-03-27 Thread Tom Lane
"Alexander B." <[EMAIL PROTECTED]> writes: > some developers were using the database, and suddenly no one could connect. Sounds like your postmaster process died. > Using PostgreSQL 8.1.4 on x86_64-pc-linux-gnu, compiled by GCC cc (GCC) > 3.3.5 (Debian 1:3.3.5-13) On Linux the most common reason

[ADMIN] Strange behavior

2007-03-27 Thread Alexander B.
Hi, some developers were using the database, and suddenly no one could connect. Just only some connections that was established before. As soon as I was advised, I noticed that no postgres process were present, just only connections and they could run queries normally. Like this: desenv:~# ps -ef

[ADMIN] Strange behavior after refreshing stored functions

2003-11-17 Thread Együd Csaba
Hi All, I have a very annoying problem at one of my customers. The phenomenon is the following: I run all function refresh scripts (create or replace function exrpssions) and trying to use them. Sometimes it seems nothing has happend to the database and the functions. They work just the same way a

[ADMIN] Strange behavior with timestamps cont.

2001-03-12 Thread Marc Wrubleski
Even more strange to me is this one below: pulver=# select date_part('epoch', timestamp 'epoch' + reltime '98529111 seconds'); date_part --- 98701911 (1 row) > > > > > > Subject: Strange behavior with timestamps

Re: [ADMIN] Strange behavior with timestamps

2001-03-12 Thread Tom Lane
Marc Wrubleski <[EMAIL PROTECTED]> writes: > pulver=# update connections set starttime = timestamp 'epoch' + reltime > '984685319'; That calculation does not do what you seem to think it does. (Hint: reltime is not simply a number of seconds.) Try just timestamp(984685319) instead.

[ADMIN] Strange behavior with timestamps

2001-03-12 Thread Marc Wrubleski
Hi all, I am witnessing something strange with my timestamps here and I am desperately hoping someone can explain or assist. If I set a time using unix timestamp, it is recalled as a different time a week different! The following example is for illustration purposes: pulver=# select version();