Hi Elke,
Only sapdb goes down. The system/other programs are ok. This problem
has being happening for some time now.
The only consistent thing about the crashes is: the last entry in the
KNLDIAGOLD is just around a minute before the crash. Always seem to be
connect requests/release such as: (this is from a previous crash not the
last one yesterday)
2004-01-29 01:00:00 0x174 19633 CONNECT Connect req. (T333,
Node:'onserluc', PID:0)
2004-01-29 01:00:00 0x174 19651 CONNECT Connection released,
T333
2004-01-29 01:00:29 0x174 19633 CONNECT Connect req. (T333,
Node:'onserluc', PID:0)
2004-01-29 01:00:29 0xDAC 19633 CONNECT Connect req. (T84,
Node:'onserluc', PID:0)
2004-01-29 01:00:29 0x174 19633 CONNECT Connect req. (T334,
Node:'onserluc', PID:0)
2004-01-29 01:00:29 0xDAC 19633 CONNECT Connect req. (T85,
Node:'onserluc', PID:0)
2004-01-29 01:00:29 0x174 19633 CONNECT Connect req. (T335,
Node:'onserluc', PID:0)
2004-01-29 01:00:50 0xDAC 19633 CONNECT Connect req. (T86,
Node:'onserluc', PID:2024)
I assume these are due to clients connecting to the database? (only my
regular client programs are running at this time. So I don't think they
are responsible.)
Could a tcp connect from some non-sapdb program to the socket kill the
database??
The KNDIAGERR only contains starting..:
2004-01-28 14:30:44 --- Starting ---
2004-01-29 01:05:46 --- Starting ---
Does this mean anything? If I remember right sapdb writes a
stack trace if it crashes. However there is none for these crashes.
The other thing I have noticed is that the antivirus gets it
updates and starts scanning the system a few seconds before the crash.
Trying a few other things currently (like disabling the
antivirus!!), will switch on the vtrace if nothing else works. (The
machine being remotely located this is the solution of last resort!!)
Thanks and regards,
Ajit
P.S. Would the knldiag from after the restart help in identifying the
problem. I have the KnlDiag /Diagold/ and knlerror log and can mail them
to you.
-----Original Message-----
From: Zabach, Elke [mailto:[EMAIL PROTECTED]
Sent: Thursday, February 19, 2004 3:04 PM
To: 'Ajit Aranha'; [EMAIL PROTECTED]
Subject: RE: Does sapdb 7.3 have any internally scheduled tasks?
Ajit Aranha wrote
There are no scheduled tasks done in the kernel,
they are not even possible.
With error log you mean knldiag?
How does it look like?
Are there any other programs ending at that time --> meaning, that the
whole
machine is somehow restarted?
If you do not have any guess, maybe you can switch on the vtrace when
leaving office
in the afternoon according to
http://sapdb.2scale.net/moin.cgi/VTrace only step 1
and in the next morning, when the instance is down (offline) and before
doing anything with it
do the steps 4..6
Elke
SAP Labs Berlin
--
MaxDB Discussion Mailing List
For list archives: http://lists.mysql.com/maxdb
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
