Really weird behaviour with 7.2 RC2

Hi,i re installed my server yesterday, upgrading mainly the os( RH 6.0 to RH7.2),Postgres 7.2b5 -> 7.2RC2 and some hardware.(raid controller)

I got like 30 000 visit / day with 2Millions hits.

the os is : RH 7.2, the cpu is an athlon 600MHz with 1.5Go SDRAM
got an Adaptec raid controller 3200S ( one array in raid 1 - 2x18Go IBM 10 000 rpm)

Top snapshot:
--------------------------------------------------------------------------------------------------------------------------------------------------------------

4:41pm up 2:43, 1 user, load average: 62,01, 58,01, 48,15
262 processes: 230 sleeping, 32 running, 0 zombie, 0 stopped
CPU states: 15,4% user, 84,5% system, 0,0% nice, 0,0% idle
Mem: 1544452K av, 492688K used, 1051764K free, 62064K shrd, 36688K buff
Swap: 2096472K av, 0K used, 2096472K free 171564K cached

PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND
3075 orion 18 0 14584 14M 2928 R 1,3 0,9 0:02 java
3073 orion 17 0 14584 14M 2928 R 1,2 0,9 0:01 java
3074 postgres 15 0 21864 21M 21156 R 1,2 1,4 0:02 postmaster
3093 orion 18 0 14584 14M 2928 R 1,2 0,9 0:01 java
1710 orion 17 0 77860 75M 2976 S 1,1 4,9 0:27 java
3000 root 17 0 1184 1184 836 R 1,1 0,0 0:06 top
3076 postgres 17 0 21192 20M 20516 R 1,1 1,3 0:01 postmaster
3095 postgres 16 0 21696 21M 20996 R 1,1 1,4 0:01 postmaster
2463 postgres 11 0 45404 44M 43936 S 1,0 2,9 0:29 postmaster
1509 orion 12 0 77812 75M 2976 R 0,9 4,9 0:28 java
1511 orion 11 0 77812 75M 2976 R 0,9 4,9 0:29 java
1585 orion 18 0 77860 75M 2976 R 0,9 4,9 0:25 java
2240 orion 13 0 77900 75M 65396 S 0,9 4,9 0:23 java
2652 orion 11 0 77900 75M 63556 S 0,9 4,9 0:13 java
2734 orion 20 0 77900 75M 55948 R 0,9 4,9 0:09 java
3031 postgres 20 0 33344 32M 32188 R 0,9 2,1 0:03 postmaster
3126 postgres 13 0 27988 27M 27300 S 0,9 1,8 0:00 postmaster
2173 orion 9 0 77900 75M 65396 S 0,8 4,9 0:18 java
2182 orion 12 0 77900 75M 65396 S 0,8 4,9 0:21 java
2184 orion 14 0 77900 75M 65396 S 0,8 4,9 0:22 java
2484 orion 10 0 77900 75M 63556 S 0,8 4,9 0:17 java
2505 orion 11 0 77900 75M 63556 S 0,8 4,9 0:11 java
2687 orion 10 0 77900 75M 55948 S 0,8 4,9 0:11 java
2886 postgres 14 0 28032 27M 26520 S 0,8 1,8 0:09 postmaster
2959 postgres 16 0 33128 32M 31984 R 0,8 2,1 0:04 postmaster
-----------------------------------------------------------------------------------------------------------------------------------------------------------

I was used to handle the load perfectly well and now its messy and i have no clue why.
The 2 things i have changed is the os: kernel 2.4.7-10 now and Postgresql 7.2RC2 instead of 7.2b5

I wonder why i have a system cpu state so high, leading to a very very high load average.
for this snapshot i started orion with no argument ( java -jar orion.jar)
As for postgres ( max backend set to the default 32):
tcpip_socket = true
shared_buffers = 16384
sort_mem = 4096
wal_buffers = 2048
wal_files = 3

Atatched with this mail is the: ps -aux | grep postgres

First question looking at this list of process, how is possible to have such many backends with max backend set to 32 ?

the log from postgres:

DEBUG: database system is shut down
DEBUG: database system was shut down at 2002-02-02 13:35:26 GMT
DEBUG: checkpoint record is at 0/71A93E4
DEBUG: redo record is at 0/71A93E4; undo record is at 0/0; shutdown TRUE
DEBUG: next transaction id: 358386; next oid: 339529
DEBUG: database system is ready
DEBUG: recycled transaction log file 0000000000000007
ERROR: Cannot insert a duplicate key into unique index eq_perso_tradeskill_pkey
ERROR: Cannot insert a duplicate key into unique index eq_perso_adv_pkey
FATAL 1: cannot open /usr/local/pgsql/data/global/1262: Too many open files in system
FATAL 1: cannot open /usr/local/pgsql/data/global/1262: Too many open files in system

If you have an idea of what is going on or how i could track down this issue..

Thx in advance
Best regards