MTS preblem on Solaris/Oracle 7.0.16

Hello,

My configuration is:	SparcStation 20/51 64Mb RAM
			Solaris 2.3 (patch SunOs 101318-49) + ODS 2.0.1
			Oracle 7.0.16.6.0
			With the procedural and distributed options

Oracle system parameters with non-default values:
  processes                = 200
  shared_pool_size         = 10485760
  control_files            = /cf/control1A1.dbf, /cf/control2A1.dbf
  db_block_buffers         = 3000
  log_archive_start        = TRUE
  log_archive_dest         = /rlf_old/log_old.dbf
  log_archive_format       = _%s.dbf
  log_buffer               = 65536
  log_checkpoint_interval  = 10000
  db_files                 = 20
  rollback_segments        = rbs1, rbs2, rbs3
  sequence_cache_hash_buckets= 10
  row_cache_cursors        = 255
  remote_os_authent        = TRUE
  db_domain                = alfa.com
  mts_service              = A1
  mts_servers              = 5
  mts_max_servers          = 25
  mts_dispatchers          = ipc,2, tcp,4
  mts_max_dispatchers      = 20
  mts_listener_address     = (ADDRESS=(PROTOCOL=ipc)(KEY=A1)), (ADDRESS=(PROTOCOL=TCP)(HOST=test)(PORT=1521))
  audit_trail              = NONE
  sort_area_retained_size  = 65536
  db_name                  = A1
  open_cursors             = 255
  optimizer_mode           = rule
  background_dump_dest     = /dest/background
  user_dump_dest           = /dest/user
  max_dump_file_size       = 10240
  core_dump_dest           = /dest/dbcore
  async_write              = TRUE
  async_read               = TRUE

and I have the following runtime situation:

Fri Aug  5							       page    1
				  USER SESSION

SERVER	  STATUS   CURRENT SESSION
--------- -------- ---------------
DEDICATED ACTIVE		 1
DEDICATED INACTIVE		 1
NONE	  INACTIVE		49
SHARED	  ACTIVE		 1


Fri Aug  5							       page    1
			      ACTIVE SHARED SERVER

   SERVER#

         5

Fri Aug  5							       page    1
				DISPATCHER LOAD

NAME			       NETWORK		BUSY STATUS


------------------------------ ---------- ---------- ----------------


D000			       ipc		   0 WAIT
D001			       ipc		   0 WAIT
D002			       tcp		 .57 WAIT
D003			       tcp		1.05 WAIT
D004			       tcp		1.02 WAIT
D005			       tcp		 .97 WAIT

Fri Aug  5							       page    1
				 LIBRARY-CACHE

NAMESPACE	      LOAD	 EXEC	  RELOAD       hit%


--------------- ---------- ---------- ---------- ----------


BODY			23	   32	       9      28.13
CLUSTER 		22	   15	       0	  0
INDEX			61	   62	       0	  0
OBJECT			 0	    0	       0	  0
PIPE			 0	    0	       0	  0
SQL AREA	    110746     441565	    3734	.85
TABLE/PROCEDURE      88516     160991	   11219       6.97
TRIGGER 		 0	    0	       0	  0

8 rows selected.

Fri Aug  5							       page    1
				    ROW-CACHE

    TOTALE     LOADED	    hit%


---------- ---------- ----------


   1497747     120176	    8.02




During normal activity the size of ROWCACHE and LIBRARYCACHE(TABLE/PROCEDURE) is continuously growing
                                                                              -----------------------

ORA-04031 out of shared memory when trying to allocate 9712 bytes

The size of shared pool buffer (9Mb !!!) seems not to be enough and we wouldn't like to increase it (if possible !!!)

Any comments or suggestions would be appreciated !!!  

Thanks a lot.
Marco  

P.S. Please email directly to tosi_at_bpb.it  

Name       : Marco Tosi
Company    : Banca Popolare di Bergamo - Credito Varesino
Address    : Piazza V.Veneto, 8 - 24122 Bergamo - Italy
email      : tosi_at_bpb.it


-----------------------------------------------------------------------