Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> Re: Oracle 7.3.2.3 "pause" problem

Re: Oracle 7.3.2.3 "pause" problem

From: Jim <jimmorgan_at_csi.com>
Date: Sat, 25 Apr 1998 12:40:01 -0500
Message-ID: <umXCvjGc9GA.326@ntawwabp.compuserve.com>


The platform is AIX. The process does nothing but updates...no inserts, no deletes. So I don't think it has anything to do with reindexing. With others' help, I think it is because Oracle is having some trouble switching redo logs...this information came from the alert log file. We're going to try increasing the number and/or size of the redo logs and see if it goes away.

--
Regards,
Jim
mpir_at_compuserve.com wrote in message <6hqu35$csl$1_at_nnrp1.dejanews.com>...
|What is your platform? If it is NT, then you could be seeing disk flushes
from
|the inserts.
|
|As to an internal ORACLE pause, it could be that the re-indexing that
occurs
|to rebalance the b-tree when the keys/leaves get out of balance.
|
|
|In article <6hn9b7$10ia$1_at_rtpnews.raleigh.ibm.com>,
| "Jim Morgan" <jimmorgan_at_csi.com> wrote:
|>
|> We have an application which does heavy-duty real-time updates from
|> TCP/IP-connected clients. We see rates of better than 30 updates per
second
|> with the tuning we have done, but we also have noticed that when the
|> database is being hit VERY hard (like when all clients were down for a
long
|> time and then they are brought online...their transactions are queued and
|> then we really start to fly!) that we get our best throughput times with
|> Oracle.
|>
|> We also notice that sometimes Oracle "takes a break" and just sits there,
|> not processing transactions, for 2 minutes or more. We cannot even start
an
|> SQL*PLUS session when this happens. So if we attempt to start an
SQL*PLUS
|> session when this happens, it just displays the first message and hangs
|> before getting the SQL*PLUS prompt. After a little while, we finally get
|> the SQL*PLUS prompt, at which time our db application begins running very
|> rapidly again.
|>
|> Anyone know what might be able to explain this behavior? We know we will
|> have some variance with the TCP/IP connected clients, but remember that
we
|> cannot even start an SQL*PLUS session when this happens, so I do not
think
|> that is the problem we are seeing.
|>
|> --
|> Regards,
|> Jim
|>
|>
|
|
|-----== Posted via Deja News, The Leader in Internet Discussion ==-----
|http://www.dejanews.com/ Now offering spam-free web-based newsreading
Received on Sat Apr 25 1998 - 12:40:01 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US