We're having an issue with Pervasive SQL 2000i SP4.

Withing the last couple of weeks, the server stops allowing incoming
Pervasive connections, while the existing connections can still remain
connected. Here's what our pvsw.log file looks like.... Does anything
in this log look like it might be an issue?

07-02-2005 00:05:11 NTMKDE 00000500 NTDBSMGR.EXE I
Resources released
07-02-2005 00:09:48 NTMKDE 0000064C NTDBSMGR.EXE I
Resources allocated
07-02-2005 03:47:19 NTMKDE 0000064C NTDBSMGR.EXE E
System Error: 620.9.33 File: J:\MAILING\DMMAIL
07-02-2005 03:48:19 NTMKDE 0000064C NTDBSMGR.EXE E
System Error: 620.9.33 File: J:\MAILING\DMMAIL
07-06-2005 09:19:26 NTMKDE 00000500 NTDBSMGR.EXE I
Resources allocated
07-06-2005 09:19:29 NTMKDE 00000500 NTDBSMGR.EXE I
The file L:\DATA\BTR\DSUSER is rolling back.
07-06-2005 09:19:38 NTMKDE 00000500 NTDBSMGR.EXE I
The file J:\PLATINUM\DSI\DSIGLOBL is rolling back.
07-07-2005 09:04:05 NTMKDE 0000066C NTDBSMGR.EXE I
Resources allocated
07-07-2005 09:04:13 NTMKDE 0000066C NTDBSMGR.EXE I
The file L:\DATA\BTR\SMTODO is rolling back.
07-07-2005 09:04:26 NTMKDE 0000066C NTDBSMGR.EXE I
The file J:\MGSYS\SCEXCEPT is rolling back.
07-07-2005 09:04:26 NTMKDE 0000066C NTDBSMGR.EXE W
The log file is corrupt.


07-07-2005 09:06:10 NTMKDE 0000066C NTDBSMGR.EXE W
The log file is corrupt.


07-09-2005 00:11:41 NTMKDE 0000066C NTDBSMGR.EXE I
Resources released
07-09-2005 00:16:15 NTMKDE 00000508 NTDBSMGR.EXE I
Resources allocated
07-09-2005 04:05:20 NTMKDE 00000508 NTDBSMGR.EXE E
System Error: 620.6.33 File: J:\MAILING\DMMAIL
07-09-2005 04:06:23 NTMKDE 00000508 NTDBSMGR.EXE E
System Error: 620.9.33 File: J:\MAILING\DMMAIL
07-09-2005 04:07:24 NTMKDE 00000508 NTDBSMGR.EXE E
System Error: 620.9.33 File: J:\MAILING\DMMAIL
07-13-2005 09:59:55 NTMKDE 00000508 NTDBSMGR.EXE I
Resources released
07-13-2005 10:00:17 NTMKDE 000004B4 NTDBSMGR.EXE I
Resources allocated
07-16-2005 00:05:10 NTMKDE 000004B4 NTDBSMGR.EXE I
Resources released
07-16-2005 00:09:46 NTMKDE 0000050C NTDBSMGR.EXE I
Resources allocated
07-16-2005 04:31:54 NTMKDE 0000050C NTDBSMGR.EXE E
System Error: 620.6.33 File: J:\MAILING\DMMAIL
07-16-2005 04:32:55 NTMKDE 0000050C NTDBSMGR.EXE E
System Error: 620.9.33 File: J:\MAILING\DMMAIL
07-18-2005 07:30:09 NTMKDE 0000061C NTDBSMGR.EXE I

I can shed some light on some of the log entries but do not believe
them to be related to the symptom that you describe.

- System Error: 620.9.33 File: J:\MAILING\DMMAIL
Comes down to a Win32 status 33,ERROR_LOCK_VIOLATION,The process
cannot access the file because another process has locked a portion of
the file.
So the database engine logs that another process has the file locked.
The users are probably getting a status 85, but it is possible that
writes are being blocked as well. Number one suspect is real-time
anti-virus software. It is best to exclude the data files from
anti-virus scans.

- The log file is corrupt.
This would be the transaction durabitlity log file that is corrupt.
Some maintenance window when the database engine is stopped go ahead
and delete the files from the "Transaction Durbility Log" directory.
By default it is pvsw\bin\mkde\log\*.* but is configurable under the
setting name above and can be anywhere.

- The file ... is rolling back.
The cause is from a system transaction (database engine write to the
data file) being interrupted. It may be more interesting to see the
messages that come before these entries. It may be related to the
System Errors (another process locking the file) that are logged.
These entries are written when the database engine unrolls the
partially writen parts of the file so the file is in a consistent
state. After this the database engine normally rolls in the
transaction durability log files to bring the data files up to the
current completed database user transaction. Since the transaction
durability log files are logged as corrupt this may or may not be
happening.

As I said, not really good stuff going on but I do not believe any of
those log entries to have anything to do with not being able to
establish additional sessions. Now if you had entries saying out of
sessions...

Leonard


On 18 Jul 2005 08:04:03 -0700, mikenospam72 (AT) hotmail (DOT) com wrote:

I'm thinking that we have someone running manual reports based on the
times (between 7AM and 10AM), and the frequency (doesn't follow a
pattern).

Anyone know a solution on how to avoid someone running manual reports
outside of locking them out (security), or purchasing a baseball bat?

Baseball bats are cheaper, but hooking up their mouse to a 220-line can
provide more shocking results. ;-)))

BTW, another possible cause of the System Error (33 errors): Running
the database application at the same time as backups. I noticed that
these occur in the wee hours of the AM -- most likely to run into
complete virus scans and backup copy jobs. If you are backing up at
these times, be sure to either get all users OUT of the database
(another use for the MickySoft Electro-Mouse), or put the database into
continuous operations mode before backup (taking it out afterwards, of
course).
Goldstar Software Inc.
Building on Btrieve(R) for the Future(SM)
Bill Bach
BillBach (AT) goldstarsoftware (DOT) com
http://www.goldstarsoftware.com
*** Chicago: Pervasive.SQL Service & Support - July 19-22 ***
*** Chicago: Pervasive DataExchange Class - July 27-29 ***

mikenospam72 (AT) hotmail (DOT) com wrote: