| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
Home -> Community -> Usenet -> c.d.o.misc -> ADO/Oracle connection problems in ONames environment
I'm having trouble pinning down a Names Server resolution issue with
multiple ADO database connections. Before the symptoms are laid out,
here's the client's environment/app particulars...
VB6/ADO 2.7/MSDAORA provider,
Oracle 9i2 in an ONAMES environment,
all clients/servers Win2k (latest service packs).
The application initially sets up two parallel ADO connections to the same database. SQLNET tracing showed immediate success on the first connection set up by the desktop. A connection was established to the Names server, the database was resolved to an IP address, and the DB connection was successfully established.
The second connection failed after identifying an existing connection to the names server (left over from the 1st DB connection?) and getting a names server timeout on the DB resolution query. The specific log entries for the failing connection read as follows...
(124) [15-DEC-2003 09:13:16:570] nigini: entry
(124) [15-DEC-2003 09:13:16:570] nigini: Count in NI global area now:
2
(124) [15-DEC-2003 09:13:16:570] nigini: Count in NI global area now:
2
(124) [15-DEC-2003 09:13:16:570] nigini: exit
(124) [15-DEC-2003 09:13:16:570] niqname: Using nnfsn2a() to build
connect descriptor for (possibly remote) database.
(124) [15-DEC-2003 09:13:16:570] nnfgrne: Going though read path
adapters
(124) [15-DEC-2003 09:13:16:570] nnfgrne: Switching to ONAMES adapter
(124) [15-DEC-2003 09:13:16:570] nnfgrne: Original name: RSTDEV1
(124) [15-DEC-2003 09:13:16:570] nngtnms_new_msg: new message, ID
6856, type 100
(124) [15-DEC-2003 09:13:16:570] nncpsrq_send_request: attempt 1,
name server with metric 0 at
(ADDRESS=(PROTOCOL=TCP)(HOST=RSTTST1)(PORT=1575)) (connection was
open)...
(124) [15-DEC-2003 09:13:16:590] nsevwait: 1 registered connection(s)
(124) [15-DEC-2003 09:13:16:590] nsevwait: 0 pre-posted event(s)
(124) [15-DEC-2003 09:13:16:590] nsevwait: waiting for transport
event (0 thru 0)...
(124) [15-DEC-2003 09:13:36:599] nsevwait: 0 newly-posted event(s)
(124) [15-DEC-2003 09:13:36:599] nsevwait: 0 posted event(s)
(124) [15-DEC-2003 09:13:36:599] nncpsrq_send_request: name server
timeout, no response in 2000 csecs
(124) [15-DEC-2003 09:13:36:599] nncpsrq_send_request: attempt 2,
name server with metric 0 at
(ADDRESS=(PROTOCOL=TCP)(HOST=RSTTST1)(PORT=1575)) (connection was
open)...
Similar tracing of multiple connections to multiple instances of SQLPLUS worksheet showed that every connection after the initial setup seemed to use cached DB Server info instead of attempting to reach the Names server for each connection.
...so, what configurations/settings/provider particulars am I messed up on here? Why can't a second connection resolve an address from the Names server? Is this a known issue betwen SQLNET and MSDAORA?
Thanks in advance... Received on Mon Dec 15 2003 - 13:55:25 CST
![]() |
![]() |