| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Usenet -> c.d.o.server -> OpenVMS 6.2, 7.1.5 PFW problem
Hi all!
We are having a very strange problem with some of our databases. We recently updated from OpenVMS 6.1 to OpenVMS 6.2 on all of our VAXEN and since have begun having strange problems in our programs that use Database links.
The problem seems to occur after all work has been done. For example:
ORACLE7-WITSK >>sqlplus
SQL*Plus: Release 3.1.3.5.1 - Production on Wed May 21 08:16:17 1997
Copyright (c) Oracle Corporation 1979, 1994. All rights reserved.
Connected to:
Oracle7 Server Release 7.1.5.2.3 - Production Release
With the distributed option
PL/SQL Release 2.1.5.2.0 - Production
Welcome to the witsk/Oracle qa database
SQL> create database link testlink using 'witskprod';
Database link created.
SQL> set timing on
SQL> select count(*) from dba_tab_columns_at_testlink;
COUNT(*)
6078
ELAPSED: 0 00:00:06.35 CPU: 0:00:00.04 BUFIO: 17 DIRIO: 41
FAULTS: 220
SQL> exit
Disconnected from Oracle7 Server Release 7.1.5.2.3 - Production Release
With the distributed option
PL/SQL Release 2.1.5.2.0 - Production
At this point the exit text from sqlplus is displayed and the process hangs for up to 30 minutes. The process is in the "PFW" state which according to our system manager indicates it is waiting for memory, but plenty of memory seems to be availaible.
$show sys
ORACLE7-WITSK >>show sys
OpenVMS V6.2-1H3 on node WITSK 21-MAY-1997 08:20:37.65 Uptime 0
21:28:16
Pid Process Name State Pri I/O CPU Page flts
Pages
000000A3 ORA_PROD_PMON HIB 6 48 0 00:00:01.64
750 726
000000A4 ORA_PROD_DBWR HIB 4 121050 0 00:01:20.15
545 604
000000A5 ORA_PROD_LGWR HIB 4 128798 0 00:01:22.23
492 513
000000A6 ORA_PROD_SMON HIB 5 1067 0 00:00:04.66 1147
1032
000000A7 ORA_PROD_RECO HIB 4 26 0 00:00:00.48
882 738
000000A8 ORA_QA_PMON HIB 6 31 0 00:00:01.41
558 671
000000A9 ORA_QA_DBWR HIB 5 17108 0 00:00:38.35
567 585
000000AA ORA_QA_LGWR HIB 5 41682 0 00:00:29.67
571 410
000000AB ORA_QA_SMON HIB 5 604 0 00:00:06.24
1310 769
000000AC ORA_QA_RECO HIB 5 17 0 00:00:00.37
825 530
000000AD TNSLSNR_QA LEF 3 3059 0 00:00:02.13
215 318
000000AE Sniper qa LEF 6 89123 0 00:04:46.36 487189 81 000000AF Sniper prod LEF 6 90094 0 00:04:55.38 608557 81 00000238 Q6Y HIB 5 313 0 00:00:00.37 204 123 00000239 Q6Y_1 HIB 5 1141 0 00:00:00.26 145 98 S 0000023A ORACLE MGR 023A PFW 4 1269 0 00:00:02.73 5799 93
$show proc/cont/id=23a
Process ORACLE MGR 023A
State PFW Working set 93
Cur/base priority 9/4 Virtual pages 14538
Current PC 800A1EB8 CPU time 0 00:00:02.73
Current PSL 00000000 Direct I/O 608
Current user SP 7EE53F60 Buffered I/O 661
PID 0000023A Page faults 5799
UIC [ORACLE7] Event flags E003D3B3
80000000
$1$DKA500:[ORACLE7.CODE7_1_5.][SQLPLUS]SQLPLUS.EXE;2 A show process/ quota for this user is shown below:
ORACLE7-WITSK >>show proc/quota
21-MAY-1997 08:25:51.16 User: ORACLE7 Process ID: 00000242
Node: WITSK Process name: "ORACLE
MGR 0242"
Process Quotas:
CPU limit: Infinite Direct I/O limit: 512 Buffered I/O byte count quota: 149808 Buffered I/O limit: 10240 Timer queue entry quota: 1024 Open file quota: 512 Paging file quota: 297920 Subprocess quota: 24 Default page fault cluster: 64 AST quota: 2046 Enqueue quota: 2048 Shared file limit: 0 Max detached processes: 0 Max active jobs: 0
And finally the account record for after the exit "FINALLY" finishes is shown:
acc /id=23a/full
ORACLE7-WITSK >>acc /id=23a /full
INTERACTIVE Process Termination
Username: ORACLE7 UIC: [ORACLE7] Account: OR7 Finish time: 21-MAY-1997 08:39:15.92 Process ID: 0000023A Start time: 21-MAY-1997 08:14:03.98 Owner ID: Elapsed time: 0 00:25:11.94 Terminal name: FTA21: Processor time: 0 00:00:02.75 Remote node addr: Priority: 4 Remote node name: Privilege <31-00>: 1411A9AD Remote ID: Privilege <63-32>: 00000008 Remote full name: Queue entry: Final status code: 00000001Queue name:
Page faults: 5799 Direct IO: 609 Page fault reads: 347 Buffered IO: 671 Peak working set: 17648 Volumes mounted: 0 Peak page file: 232592 Images executed: 9
We have an open TAR with oracle, they have suggested and we have upped the page file quota, byte limit, npagedyne, we have also increased the memory on the machine since the problem began. I have talked to world wide support so many times on this TAR that I now have it memorized. They are working this issue hard but seem to be stumped. I can find no issues with 6.2 and database links on the MetaLink page.
Does any one have any ideas?
Thanks in advance,
John Hough Received on Wed May 21 1997 - 00:00:00 CDT
 |
 |