Message-Id: <10697.123400@fatcity.com> From: Alex Hillman Date: Fri, 1 Dec 2000 10:20:07 -0500 Subject: RE: Frequency of Log Switches This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01C05BAA.2FEDD400 Content-Type: text/plain; charset="iso-8859-1" Let say you ship your archivelogs to different machine and let say that your disks where you have your redo logs are annihilated (explosion, fire, etc.). You will lose all transaction activity in current redo log - doesn't matter how often you are doing checkpoints. So the only way to be sure that you will not loose more than 15 min. of data is to make log switch not later then 15 min. after previous switch to archive and ship to different destination. Alex Hillman -----Original Message----- From: Michael Ray [mailto:Michael.Ray@trw.com] Sent: Friday, December 01, 2000 8:51 AM To: Multiple recipients of list ORACLE-L Subject: RE: Frequency of Log Switches It seems people are arguing over semantics, and the original poster was asking the wrong question. All he wanted was to not lose more than 15 minutes of data if he crashed. This is controlled by checkpoints, NOT log switches. Setting the parameters Kieran mentioned appropriately will accomplish that (along with FAST_START_IO_TARGET if using 8i). The INTERVAL parameter specifies the maximum number of OS blocks you wish to be read in case of instance recovery. The TIMEOUT one specifies the maximum seconds of blocks you wish to read which obviously is much more ambiguous and why I believe Satar suggested staying away from it. If you are pre-8i these parameters have different meanings. If you wish to force a log switch, ALTER SYSTEM SWITCH LOGFILE; but that is only going to increase IO on your system and gain you nothing. Michael Ray Oracle DBA TRW, Marshall, IL 217-826-3011 x2438 >>> ASAURY@recherche.loreal.com 11/30/00 04:55AM >>> I beg to differ : I am running 8.1.6 on NT with log_checkpoint_timeout = 300 and can have as much as 40 minutes between log switches. I suppose that forcing a checkpoint does not forces a log switch (whereas the log switch forces the checkpoint). Anybody having an idea on how to force a log switch ? Alain. -----Message d'origine----- Steven, try changing the initialization parameter LOG_CHECKPOINT_TIMEOUT to be 900 and set the LOG_CHECKPOINT_INTERVAL to be a very high value (greater than the size of the redo logs), this will cause log switches every 15 minutes or when the log file is full. Both of these values can be dynamically altered using ALTER SYSTEM. Regards, Kieran Murray Norkom Technologies, 43 Upper Mount Street, Dublin 2, Ireland Tel:- +353 86 8117916 -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Michael Ray INET: Michael.Ray@trw.com Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051 San Diego, California -- Public Internet access / Mailing Lists -------------------------------------------------------------------- To REMOVE yourself from this mailing list, send an E-Mail message to: ListGuru@fatcity.com (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). ------_=_NextPart_001_01C05BAA.2FEDD400 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable RE: Frequency of Log Switches

Let say you ship your archivelogs to different = machine and let say that your disks where you have your redo logs are = annihilated (explosion, fire, etc.).

You will lose all transaction activity in current = redo log - doesn't matter how often you are doing checkpoints. So the = only way to be sure that you will not loose more than 15 min. of data = is to make log switch not later then 15 min. after previous switch to = archive and ship to different destination.

Alex Hillman

-----Original Message-----
From: Michael Ray [mailto:Michael.Ray@trw.com]
Sent: Friday, December 01, 2000 8:51 AM
To: Multiple recipients of list ORACLE-L
Subject: RE: Frequency of Log Switches


It seems people are arguing over semantics, and the = original poster was asking the wrong question. All he wanted was to not = lose more than 15 minutes of data if he crashed. This is controlled by = checkpoints, NOT log switches.

Setting the parameters Kieran mentioned appropriately = will accomplish that (along with FAST_START_IO_TARGET if using 8i). The = INTERVAL parameter specifies the maximum number of OS blocks you wish = to be read in case of instance recovery. The TIMEOUT one specifies the = maximum seconds of blocks you wish to read which obviously is much more = ambiguous and why I believe Satar suggested staying away from it. If = you are pre-8i these parameters have different meanings.

If you wish to force a log switch, ALTER SYSTEM = SWITCH LOGFILE; but that is only going to increase IO on your system = and gain you nothing.

Michael Ray
Oracle DBA
TRW, Marshall, IL
217-826-3011 x2438

>>> ASAURY@recherche.loreal.com 11/30/00 = 04:55AM >>>
I beg to differ : I am running 8.1.6 on NT with = log_checkpoint_timeout =3D 300
and can have as much as 40 minutes between log = switches.

I suppose that forcing a checkpoint does not forces a = log switch (whereas
the log switch forces the checkpoint).

Anybody having an idea on how to force a log switch = ?
Alain.

        -----Message d'origine-----

        Steven,

        try = changing the initialization parameter LOG_CHECKPOINT_TIMEOUT to
be 900
        and set = the LOG_CHECKPOINT_INTERVAL to be a very high value (greater
than
        the size = of the redo logs),
        this will = cause log switches every 15 minutes or when the log file
is full.
        Both of = these values can
        be = dynamically altered using ALTER SYSTEM.

        Regards,
        Kieran = Murray
        Norkom = Technologies,
        43 Upper = Mount Street, Dublin 2, Ireland
        Tel:- = +353 86 8117916

--
Please see the official ORACLE-L FAQ: http://www.orafaq.com
--
Author: Michael Ray
  INET: Michael.Ray@trw.com

Fat City Network Services    -- (858) = 538-5051  FAX: (858) 538-5051
San Diego, = California        -- Public Internet = access / Mailing Lists
---------------------------------------------------------------= -----
To REMOVE yourself from this mailing list, send an = E-Mail message
to: ListGuru@fatcity.com (note EXACT spelling of = 'ListGuru') and in
the message BODY, include a line containing: UNSUB = ORACLE-L
(or the name of mailing list you want to be removed =