Perrenes bazu s odnogo servera na drugoj.

solaris SunOS 5.8 Generic Patch
oracle 8.1.7.4.
maschine E15K 8 Processors

ransche tnsping schol zu 10-100ms
teper 20000-70000.

Perkonfiguriroval na druguju Netzwerkkartu- effect totzhe.
listener.ora:

listener_XXXX =
(ADDRESS_LIST =
(ADDRESS=(PROTOCOL=tcp)(HOST=xxxx-pr1)(PORT=7022)(queuesize=32))
(ADDRESS=(PROTOCOL=tcp)(HOST=xxxx-pr2)(PORT=7222)(queuesize=32))
)

SID_LIST_LISTENER_XXXX =
(SID_LIST =
(SID_DESC =
(SDU = 1460)(TDU = 1460)
(GLOBAL_DBNAME = XXXX)
(ORACLE_HOME = /opt/oracle/product/817)
(SID_NAME = XXXX)
)
)
Pervyj Interface dlya UNIX, vtoroj - dlya CITRYX, ostlasy pri perenose bazy bez izmenenij.

Chertovschina kakaja-to:(((

vse nastrojeno kak nado.

oracle@borys2:/opt/oracle>netstat -i
Name Mtu Net/Dest Address Ipkts Ierrs Opkts Oerrs Collis Queue
lo0 8232 loopback localhost 16650612 0 16650612 0 0 0
ce0 1500 borys2-hb2 borys2-hb2 339785151 0 5909144 0 0 0
ce2 1500 borys2-hb3 borys2-hb3 121959316 0 5909433 0 0 0
ce3 1500 borys2-e5-p borys2-e5-p 283616399 0 0 0 0 0
dman0 1500 10.1.1.0 10.1.1.3 69389 0 563260 2 0 0
ge0 1500 borys2-g0-p borys2-g0-p 2600496210 0 2838015456 0 0 0
ge1 1500 borys2-g1-p borys2-g1-p 6043098 0 3 0 0 0
qfe0 1500 borys2-e0 borys2-e0 49955946 0 32417845 0 0 0
qfe1 1500 borys2-hb1 borys2-hb1 365126696 50780 5909148 0 0 0
qfe2 1500 borys2-e2-p borys2-e2-p 994815701 0 728023078 0 0 0

Listener sluschajet na ge0- Interface

tut vot jescho client trace dlya DB

....................
[06-MAY-2003 11:54:45] nscon: recving a packet
[06-MAY-2003 11:54:45] nsprecv: entry
[06-MAY-2003 11:54:45] nsbal: entry
[06-MAY-2003 11:54:45] nsbgetfl: entry
[06-MAY-2003 11:54:45] nsbgetfl: normal exit
[06-MAY-2003 11:54:45] nsmal: entry
[06-MAY-2003 11:54:45] nsmal: 44 bytes at 0x316580
[06-MAY-2003 11:54:45] nsmal: normal exit
[06-MAY-2003 11:54:45] nsbal: normal exit
[06-MAY-2003 11:54:45] nsprecv: reading from transport...
[06-MAY-2003 11:54:45] nttrd: entry
[06-MAY-2003 11:55:08] nttrd: socket 192 had bytes read=8
[06-MAY-2003 11:55:08] nttrd: exit
[06-MAY-2003 11:55:08] nsprecv: 8 bytes from transport
[06-MAY-2003 11:55:08] nsprecv: tlen=8, plen=8, type=11
[06-MAY-2003 11:55:08] nsprecv: packet dump
[06-MAY-2003 11:55:08] nsprecv: 00 08 00 00 0B 00 00 00 |........|
[06-MAY-2003 11:55:08] nsprecv: normal exit
[06-MAY-2003 11:55:08] nscon: got NSPTRS packet
[06-MAY-2003 11:55:08] nscon: no connect data
[06-MAY-2003 11:55:08] nscon: exit (0)
[06-MAY-2003 11:55:08] nsdo: nsctxrnk=0
[06-MAY-2003 11:55:08] nsdo: normal exit
.........................

25 Secund dlilsya connect s sqlplus:(((

mgnovenno.
Na etom zhe interface, no s drugim Virtual-Ip i na drugom Portu sluschajet Listener dlya vtoroj DB. tnsping na etu druguju DB dlitsya 20-30 ms. Obe DB sravnimy po kollichestvu connectov.
:-((((

truss na Clienta:
...............
1087/1: 0.0017 sigaction(SIGPIPE, 0xFFBE9D60, 0xFFBE9E8C) = 0
1087/1: 0.0016 getpid() = 1087 [6888]
1087/1: 0.0009 write(8, "\0D2\0\001\0\0\001 601 ,".., 210) = 210
1087/1: read(8, 0x000EDF76, 1476) (sleeping...)
1087/2: signotifywait() (sleeping...)
1087/3: lwp_cond_wait(0xFEBF5548, 0xFEBF5558, 0xFEBEEDB0) (sleeping...)
1087/4: door_return(0x00000000, 0, 0x00000000, 0) (sleeping...)
1087/1: 33.7050 read(8, "\0\b\0\0\v\0\0\0", 1476) = 8
1087/1: 0.0012 write(8, "\0D2\0\001\0\0\001 601 ,".., 210) = 210
1087/1: 0.0004 read(8, "\0 \0\002\0\0\001 6\f01".., 1476) = 32
1087/1: 0.0008 brk(0x000EF990) = 0
1087/1: 0.0004 brk(0x000F1990) = 0
1087/1: 0.0011 sigaction(SIGTSTP, 0xFFBEC470, 0xFFBEC59C) = 0
................

33,7 secundy pytajetsya chet prochitat?:(

vse!!!
reshili problemu.
Unixoviki krivo /opt/oracle smontirovali!!!!

...read/write/setuid/mincache=closesync/delaylog/largefiles/usw...

ubrali option mincache=closesync, i teper tnsping prohodit za 10 ms.
Spasibo vsem za vnimanije!!!
)))
)))

Vo blin!!!
Meloch, a gora s plech!!!!
)