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

Home -> Community -> Mailing Lists -> Oracle-L -> RE: duplexing redos and archives

RE: duplexing redos and archives

From: Fink, Dan <Dan.Fink_at_mdx.com>
Date: Tue, 06 Aug 2002 15:29:51 -0800
Message-ID: <F001.004ACF69.20020806152951@fatcity.com>


I've never personally had to deal with #2, nor have I heard of anyone experiencing this particular situation. In fact, I hope never to be able to say I had to deal with a logically corrupt redo log. I unfortunately have dealt with corrupted archive logs and the long hours spent rebuilding a db have made me very, very protective of my redo. Compare the cost of 1g of space in a 500g array against an unrecoverable db.

However, if the redo logs are written using different hardware paths, it is possible to have 1 with an untrapped corruption while the other is okay. Again this is theory...

I recently encountered a situation where duplexing saved my behind. On a non-prod system, we reconfigured the disks. Prior to shutdown, I moved some the redo logs off of the disks that were going away. A miscommunication resulted in the SA starting the dbs before I had a chance to rename the redo logs. Since I was duplexing (they are mirrored at the hardware level as well), the database started (tossed a couple of errors in the alert log) without problem. When I finally realized what had happened, I simply dropped and recreated the log groups (needed to resize them anyway...) with no additional downtime.

Dan

-----Original Message-----
Sent: Tuesday, August 06, 2002 4:29 PM
To: Multiple recipients of list ORACLE-L

Has anyone ever heard of #2 ever occuring? Even if possible I am not aware of anyone presenting a clear case of this actually happening.

Ethan Post
(972) 577-6552
Ethan.Post_at_ps.net
perotdba (AIM), epost1 (Yahoo)


-----Original Message-----
Sent: Tuesday, August 06, 2002 4:08 PM
To: Multiple recipients of list ORACLE-L

I think the logical argument is on your side as well. The Oracle kernel has some code path that allows it to write different things into supposedly "identical" redo logs, to account for device glitches, etc. But this works only if you use the *Oracle* capability to multiplex redo.

Cary Millsap
Hotsos Enterprises, Ltd.
http://www.hotsos.com

Upcoming events:
- Hotsos Clinic, Oct 1-3 San Francisco, Oct 15-17 Dallas, Dec 9-11 Honolulu
- 2003 Hotsos Symposium on OracleR System Performance, Feb 9-12 Dallas - Next event: NCOAUG Training Day, Aug 16 Chicago

-----Original Message-----
Sent: Tuesday, August 06, 2002 1:44 PM
To: Multiple recipients of list ORACLE-L

This may be 'old school' thinking, but I prefer to duplex my redo logs. Why?
Because there are some things that mirroring cannot protect you against. 1. The deletion of a redo log through human error (been there, recovered from it...)
2. The logical corruption of a redo log (never experienced it myself, but it
is theortically possible)
3. Oracle's fault tolerance regarding the ability to function as long as 1
redo log in a group is accessible

Considering the critical nature of the redo logs (lose one and your recovery
is done...and perhaps your job) and their small size in relation to overall
databse size, duplexing in a group makes sense.

This is one of those issues where valid arguements exist on both sides. In
some ways, the logical arguements are on the side of not duplexing redo logs
that are mirrored. However, when you consider that the loss or corruption of
a single redo log can cause the database to be unrecoverable is enough to
justify the cost of duplexing.

Just my humble (old school) opinion

Dan Fink

-----Original Message-----
Sent: Tuesday, August 06, 2002 12:13 PM
To: Multiple recipients of list ORACLE-L

Currently we are duplexing our redo logs and archive logs through oracle. We do this at the oracle level because we are using mirroring (raid0?) and if we lose a redo log or archive file the database will not be affected.

We will be moving to raid5 arrays in the future and are wondering whether we still need to do the duplexing. Since the arrays are made up of data disks plus a parity disk plus a hot swap do we need to worry about losing a redo log? Our sys admins are telling us we don't and I tend to agree with them but we want to get other opinions on the subject.

How are shops handling this who are using raid5?

Thanks.
Joe
--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author: Joe Armstrong-Champ
  INET: Joseph.Armstrong-Champ_at_tufts.edu

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_at_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).
--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author: Fink, Dan
  INET: Dan.Fink_at_mdx.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_at_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).

--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author: Cary Millsap
  INET: cary.millsap_at_hotsos.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_at_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).
--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author: Post, Ethan
  INET: Ethan.Post_at_ps.net
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_at_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).
--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author: Fink, Dan
  INET: Dan.Fink_at_mdx.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_at_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). Received on Tue Aug 06 2002 - 18:29:51 CDT

Original text of this message

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