Couple of quick things,
do you have to restart the whole machine to get oracle working again, or do
you just start the service again?
as a quick idea for the moment, if it would not cause any inconvenience at
the time, would it not be an idea to schedule a batch file from your machine
or another to run a batch file, similar to the folloing at 1am or so:
SC \\servername STOP service
SC \\servername START service
you will forgive, but it has been a while since i have used this, and i
cannot remember the correct syntax for the command.
i believe SC is in the reskit.
(the idea of this being that if you stop and start the service yourself, you
know it has been made "fresh" every night :)
Also i just thought, the only other reason i can think of are out of
date/badly configured/corrupted oracle clients on one or more of your client
machines.
This test could actually prove this, it might be an idea to log who is using
the oracle server at the time, (ie check the log before you start the
service again), whoever logged in last, uninstall and reinstall the oracle
client on that machine.
i do believe i have seen this before.
thats all i can think of for the minute. let me know.
thanks
anthony
----- Original Message -----
From: "Doug Lott" <[EMAIL PROTECTED]>
To: "NT System Admin Issues" <[EMAIL PROTECTED]>
Sent: Tuesday, September 04, 2001 4:25 PM
Subject: Oracle just Dies on NT 4 sp6a
> We have eight NT4 sp6a servers running Oracle where Oracle just dies.
> That is Tivoli reports the srvice for Oracle is stopped. When checked it
> is stoppped and there are no errors in any eventlog nor in any Oracle
> logs. The Oracle DBA and myself are stumped. We cannot find any errors
> or records of any recent changes to the systems. These problems were
> noticed about In the past three months or so ago.
>
> Four serveres are running Oracle 7.3.4 and the other four are running
> Oracle 8.1.6. Each week at least one of them drops out (the Oracle
> service that is). A server may run for a month or two before the Oracle
> service stops. Some servers are in Dallas and some are in NY. All have a
> similare frequency of failure and it happens at various times of day.
>
> Has anyone experienced this? Or have any ideas where we can start looking
> or what we can monitor? Technet and Oracle knowledge bases do not seem to
> record any thing similar.
>
> Thanks in advance
>
> Doug.
>
> http://www.sunbelt-software.com/ntsysadmin_list_charter.htm
>
>
http://www.sunbelt-software.com/ntsysadmin_list_charter.htm
