Redo Log Buffer 32.8 M, looks great?

Similar Questions

• Redo log buffer question

  Hi master,
  
  This seems to be very basic, but I would like to know internal process.
  
  We know all that LGWR writes redo entries to redo logs on the disk online. on validation SCN is generated and the tag to the transaction. and LGWR writes this to online redo log files.
  
  but my question is, how these redo entries just redo log buffer? Look at all the necessary data are read from the cache of the server process buffers. It is modified it and committed. DBWR wrote this in files of data, but at what time, what process writes this committed transaction (I think again entry) in the log buffers cache?
  
  LGWR do that? What exactly happens internally?
  
  If you can please focus you some light on internals, I will be grateful...
  
  
  Thanks and greetings
  VD

  Vikrant,
  I will write less coz used pda. In general, this happens
  1. a calculation because how much space is required in the log buffer.
  2 server process acquires redo copy latch to mention some reco will be responsible.

  Redo allocation latch is used to allocate space.

  Redo allocation latch is provided after the space is released.

  Redo copy latch is used copy redo contained in the log buffer.

  Redo copy lock is released

  HTH
  Aman

• REDO LOG BUFFER DOUBT

  Hi EXPERTS,
  
  IM on 11 G R2, RHEL 5 my question that when the data changes in the database buffer cache how immediately it is copied in the redo log buffer. I mean what process the copy. I read it in the oracle documentation only when waiting for event [log file command (private stream flush incomplete)] LGWR waits DBWR again complete flushing of buffers IMU in the recovery log buffer. Completion of DBWR LGWR can then finish writing the log in process and then move the log files.
  

• relationship between redo log buffer, journal of redo and undo tablespace files

  What is the relationship between the redo log buffer, redo log files and undo tablespace?
  
  what I understand is
  
  redo log buffer is the logical area where all the news of recovery were stored until they are transferred by LGWR bank roll forward log online... but y at - it report any with the undo tablespace with these two?
  
  Please correct me if I'm wrong
  
  Thanks in advance

  redo log buffer is the logical area where all the news of recovery were stored until they are transferred by LGWR bank roll forward log online... but y at - it report any with cancellations

  tablespace with these two?

  There is a link between files redo log and buffer, but the undo tablespace is something else entirely.

  spend it here with this links

  REDO LOG FILES
  http://www.DBA-Oracle.com/concepts/redo_log_files.htm

  BUFFER REDOLOG
  http://www.DBA-Oracle.com/concepts/redo_log_buffer_concepts.htm

  UNDO TABLESPACE
  Undo tablespace to cancel files to undo or roll back uncommitted changes pray the database.

  Hope you understood.

• Question about the size of the redo log buffer

  Hello
  
  I am a student in Oracle and the book I use says that having a bigger than the buffer log by default, the size is a bad idea.
  
  It sets out the reasons for this are:
  
  >
  The problem is that when a statement COMMIT is issued, part of the batch validation means to write the contents of the buffer log for redo log to disk files. This entry occurs in real time, and if it is in progress, the session that issued the VALIDATION will be suspended.
  >
  
  I understand that if the redo log buffer is too large, memory is lost and in some cases could result in disk i/o.
  
  What I'm not clear on is, the book makes it sound as if a log buffer would cause additional or IO work. I would have thought that the amount of work or IO would be substantially the same (if not identical) because writing the buffer log for redo log files is based on the postings show and not the size of the buffer itself (or its satiety).
  
  Description of the book is misleading, or did I miss something important to have a larger than necessary log buffer?
  
  Thank you for your help,
  
  John.
  
  Published by: 440bx - 11 GR 2 on August 1st, 2010 09:05 - edited for formatting of the citation

  A commit evacuates everything that in the buffer redolog for redo log files.
  A redo log buffer contains the modified data.
  But this is not just commit who empty the redolog buffer to restore the log files.
  LGWR active every time that:
  (1) a validation occurs
  (2) when the redo log is 1/3 full
  (3) every 3 seconds
  It is not always necessary that this redolog file will contain validated data.
  If there is no commit after 3 seconds, redologfile would be bound to contain uncommitted data.

  Best,
  Wissem

• Redo Log Buffer - Online redo - archived journal - LGWR

  Hello
  
  It is specified in this document that redo entries is set to all changes (LMD), the database Redo Log Buffer (my question is when does? after validation?).
  
  And, in this connection, they say engaged and committed changes in the section again. But when you scroll down to the section Commit on the same link, after a commit log writer process (LGWR) will write journal entries of the SGA (log buffer) forward in the line redo log file.
  
  I wonder how? When I tested it on my database table that holds the record of > 300 000, recovery, and subsequently logs archive logs is generated (I did not commit the DML yet).
  
  Now my question is, how far is the whole of the archivelogs will be useful at the time of recovery, if I restore the changes? If Oracle doesn't know these archivelogs, why it should generate?
  
  And what section in the second link is correct?
  
  Help, please.
  
  Thank you
  Aswin.

  Aswin,

  Undo segments are ordinary data segments. Data segments are protected by roll forward. Temporary segments are not protected restore in progress.
  If in the case of an accident, when Oracle must retrieve a rollback segment, it will use again to recover this segment.
  In crash recovery, the first thing that will happen is rollback (Undo uncommitted changes), followed by rollforward ("readback" of committed changes).

  HTH

  -----------------
  Sybrand Bakker
  Senior Oracle DBA

• That redo log files waiting?

  Hello Experts,

  I read articles on the log redo and undo segment files. I was wondering something very simple. That redo log files waiting in there? It stores the sql statements?

  Lets say that my update statement to modify 800 blocks of data. A unique single update statement can modify different data 800 right blocks? Yes, it may be true. I think that these data blocks can not hold buffers to the log to roll forward, right? I mean I know exactly what to do redo log buffer and redo log file. And I know that the task of backgrounding LGWR. But, I wonder if she she holds the data blocks? It is not supposed to hold data like cache buffer blocks, right?

  My second question is, rollback isn't effect to restore the newspaper to the right buffer? Because it does not need log buffer for effect do it again. Conversely, the restoration; statement is included in the restore log buffer by progression when someone isse, am I right?

  As far as I know, rollback interact directly with UNDO TABLESPACE?

  I hope that I have to express myself clearly.

  Thanks in advance.

  Here's my question:

  My second question is, rollback isn't effect to restore the newspaper to the right buffer? Because it does not need log buffer for effect do it again. Conversely, the restoration; statement is included in the restore log buffer by progression when someone isse, am I right?

  As far as I know, rollback interact directly with UNDO TABLESPACE?

  Yes, where else would the undo data come from? Undo tablespace contains the Undo segments that contain the Undo data required for the restoration of your transaction.

  I can say that rollback does not alter the data of the log buffer rede to the past. In other words, change vectors will be remain the same before restoration. Conversely, rollback command is also recorded in the log file of restoration by progression. As the name, all orders are saved in the REDO LOGS.

  I hope that I am wrong so far?

  Not sure why you even the buffer log roll forward for Rollback? This is the reason why I asked you it was for, where occurs the dose the cancellation? And the answer for this is that it happens in the buffer cache. Before you worry about the drivers of change, you must understand that it is not serious what contains where as long as there is no transaction recorded in the operating of the Undo segment table. If the operating table indicates that the transaction is longer there, there must be a cancellation of the transaction. Vectors of change are saved in the file log roll forward, while the restore happens on blocks of data stored in the file "data" undo blocks stored in the undo file "data".

  At the same time I read an article about redo and undo. In this article process transaction is explained. Here is the link http://pavandba.files.wordpress.com/2009/11/undo_redo1.pdf

  I found some interesting information in this article as follows.

  It is worth noting that during the restore process, recovery logs never participate. The only time where redo logs are read is retrieving and archiving. This is the concept of tuning key: redo logs are written on. Oracle does not read during normal processing. As long as you have sufficient devices so that when the ARC is reading a file, LGWR's writing to a different device, then there no contention for redo logs.

  If redo logs are never involved in the restoration process, how is it Oracle will then know the order of the transaction? As far as I know it is only written in redo logs.

  I have thoughts very amazed to Aman.

  Why you ask?

  Now, before giving a response, I say two things. One, I know Pavan and he is a regular contributor to this forum and on several other forums Facebook and two, with all due respect to him, a little advice for you, when you try to understand a concept, to stick to the Oracle documentation and do not read and merge articles/blog-posts from the web. Everone, which publishes on the web, has their own way to express things and many times, the context of the writing makes it more confusing things. Maybe we can erase the doubts that you can get after reading the various search results on the web.

  Redo logs used for the restoration, not to restore. The reason is the redo log files are applied in sequential order, and this is not the case when we look for the restoration. A restore is required to do for a few blocks away. Basically, what happens in a restoration, is that the records of cancellation required for a block of data are sought in the reverse order of their creation. The entry of the transaction is in the slot ITL of the block of data that point to the necessary undo bytes Address (UBA) using which oracle also knows what that undo the blocks would be necessary for the restoration of your transaction. As soon as the blocks of data will be cancelled, the ITL slots would be cleared as well.

  In addition, you must remember, until the transaction is not qualified as finished, using either a commit or a rollback, the cancellation of this data would remain intact. The reason for this is that oracle would ensure that undo data would be available to make the cancellation of the transaction. The reason why Undo data are also recorded in the journals of recovery is to ensure that in the event of the loss of the cancellation of the data file, retrieving them would be possible. Because it would also require changes that's happened on the blocks cancel, restore the vectors change associated with blocks of cancellation are also saved in the buffer log roll forward and, in the redo log files.

  HTH

  Aman...

• Redo log entries to the online redo log files

  Hello

  I'm a little confused about the redo log entries.

  I learned that LGWR writes redo log entries in the SGA, in redo log files in the following situation:

  -3 seconds after the last validation

  -redo log buffer has 1 MB

  -redo log buffer data 1/3

  When I take a transaction and if I change the wrong data, I have to cancel this transaction BUT the right now,

  What happens if the LGWR writes redo log disk entries?

  Is the bad transaction committed at all times?

  2791011 wrote:

  Hello

  I'm a little confused about the redo log entries.

  I learned that LGWR writes redo log entries in the SGA, in redo log files in the following situation:

  -3 seconds after the last validation

  -redo log buffer has 1 MB

  -redo log buffer data 1/3

  When I take a transaction and if I change erroneous data, I have to cancel this transaction at this time, BUT

  What happens if the LGWR writes redo log disk entries?

  Is the bad transaction committed at all times?

  DML changes are "permanent" after that VALIDATION is issued.

  When any COMMIT is issued, then changes can be undone.

• How do to the size of the log buffer reduce with huge Pages Linux

  Data sheet:

  Database: Oracle Standard Edition 11.2.0.4

  OS: Oracle Linux 6.5

  Processor: AMD Opteron

  Sockets: 2

  Carrots / power outlet: 16

  MEM: 252 GB

  Current SGA: GB 122 automatic shared memory (EAMA) management

  Special configuration: Linux huge Pages for 190 GB of memory with the page size of 2 MB.

  Special configuration II: The use of the LUKS encryption to all drives.

  Question:

  1. How can I reduce the size of the log buffer? Currently, it appears that 208 MB. I tried to use log_buffer, and it does not change a thing. I checked the granule size is 256 MB with the current size of the SGA.

  Reason to reduce:

  With the largest size of log buffer the file parallel write newspaper and the synchronization log file is averaged over 45 ms most of the time because she has dumped a lot of stuff at the same time.

  Post edited by: CsharpBsharp

  You have 32 processors and 252 GB of memory, so 168 private discussions, so 45 MB as the public threads size is not excessive.  My example came from a machine running Oracle 32-bit (as indicated by the size of 64 KB of threads private relative to the size of 128 KB of your son) private with (I think) 4 GB of RAM and 2 CPUs - so a much smaller scale.

  Your instance was almost inactive in the meantime so I'd probably look outside the Oracle - but verification of the OS Stats may be informative (is something outside the Oracle using a lot of CPU) and I would like to ask you a few questions about encrypted filesystems (LUKS).  It is always possible that there is an accounting error - I remember a version of Oracle who report sometimes in milliseconds while claiming to centisecondes - so I'll try to find a way to validate the log file parallel write time.

  Check the forum for any stats activity all over again (there are many in your version)

  Check the histogram to wait for additional log file writes event (and journal of file synchronization - the lack of EPA top 5 looks odd given the appearance of the LFPW and and the number of transactions and redo generated.)

  Check the log file to track writer for reports of slow writes

  Try to create a controlled test that might show whether or not write reported time is to trust (if you can safely repeat the same operation with extended follow-up enabled for the author of newspaper which would be a very good test of 30 minutes).

  My first impression (which would of course be carefully check) is that the numbers should not be approved.

  Concerning

  Jonathan Lewis

• bottleneck during the passage of the redo log files.

  Hi all
  
  I'm using Oracle 11.2.0.3.
  
  The enforcement team has indicated that they are facing slow at some point.
  
  I followed the database and I found that at some passage of redo log files (not always), I am facing a slow at the application level.
  
  I have 2 son since my database is CARS, each thread has 3 groups of multiplexed redo logs the FIU, with size of 300 MB each.
  
  Is it possible to optimize the switch of the redo log files? knowing that my database is running in ARCHIVELOG mode.
  
  Kind regards

  Hello

  Yes, Oracle recommends 1 validation by 30 calls from users or less. Of course, every database is different, so this rule cannot be taken too literally, but in your case, this rule seems to apply. In any case State, 900 undertakes seconds it looks like a very large number and the need for a high number of transactions should be questioned. You should talk to your analysts/application management/enterprise architect if warranted - that is to say if the application does in fact almost 2 000 business transactions per second.

  What about DB CPU: here is a link to a blog that I wrote on this subject, it should help to better understand this number:

  http://Savvinov.com/2012/04/06/AWR-reports-interpreting-CPU-usage/

  But briefly: DB processor isn't a real event, it is simply an indication that the sessions are on CPU (or waiting for CPU) rather than wait on i/o requests or events in the database. It is not necessarily a bad thing, because the database must perform tasks and he cannot do without loading the CPU. It may indicate a problem in two cases: when the CPU usage is close to the limit of the host (OS stats section indicates that you are very far from there) or when the CPU is a % of DB time - in the latter case, this could mean that you are making too many logical reads due to inefficient plans or analysis too. In any case, this does not apply to you, because 20 percent is not very high a number.

  Other items in the list of the top 5 deserve attention, too - gc buffer busy acquire, gc current block busy, enq: TX - line lock conflict.

  To summarize, your database under a lot of stress - whether it is the legitimate workload, and if this is the case, you may need to upgrade your hardware later. There is chance that it isn't - for example a high number of runs may indicate that rather than to bulk operations database code using PL/SQL loops, which is a big performance killer. Check "Top SQL by executions" about whether or not this is the case.

  Good luck!

  Best regards
  Nikolai

• Waiting for redo log

  Hello
  in 10g R2 what are / can be the causes of waiting for redo log?
  
  Thank you.

  example of waiting for recovery logs causes may be:
  -does not redo groups (remember to add one more, if necessary)
  -low rate of I/O
  -Redo large log buffer (take a long time for the dump file data) - refers to the slow storage
  -application hurt writes, making many validations

  Run the Advisor of the ADDM, and see the recommendations. Look at the first 5 albums events. Maybe it's not relevant to DBTime?

• How do I know if there's bottleneck in the redo logs?

  Hi all

  EBS R12.2

  11 GR 2

  OL 6.5

  We have two groups of log of size 1 GB.

  and oyour archiving logs generates 13 newspapers every hour.

  -rw - r-. 1 oraprod s/n 956896768 Jan 25 14:00 1_3372_898613761.dbf

  -rw - r-. 1 oraprod s/n 1004083200 Jan 25 14:03 1_3373_898613761.dbf

  -rw - r-. 1 oraprod s/n 928530432 Jan 25 14:10 1_3374_898613761.dbf

  -rw - r-. 1 oraprod s/n 928728576 Jan 25 14:12 1_3375_898613761.dbf

  -rw - r-. 1 oraprod s/n 967805952 Jan 25 14:20 1_3376_898613761.dbf

  -rw - r-. 1 oraprod s/n 916065792 Jan 25 14:22 1_3377_898613761.dbf

  -rw - r-. 1 oraprod s/n 951790592 Jan 25 14:30 1_3378_898613761.dbf

  -rw - r-. 1 oraprod s/n 978358272 Jan 25 14:32 1_3379_898613761.dbf

  -rw - r-. 1 oraprod s/n 974519808 Jan 25 14:40 1_3380_898613761.dbf

  -rw - r-. 1 oraprod s/n 960421376 Jan 25 14:42 1_3381_898613761.dbf

  -rw - r-. 1 oraprod s/n 917438976 Jan 25 14:49 1_3382_898613761.dbf

  -rw - r-. 1 oraprod s/n 920794624 Jan 25 14:51 1_3383_898613761.dbf

  -rw - r-. 1 oraprod s/n 920704000 Jan 25 14:59 1_3384_898613761.dbf

  I got this alert saves messages:

  Mon Jan 25 15:08:37 2016

  Filled checkpoint up to RBA [0xd3b.2.10], RCS: 5978324588151

  Mon Jan 25 15:10:57 2016

  Thread 1 cannot allot of new newspapers, sequence 3388

  Private stream flush is not complete

  Currently journal # 1, seq # 3387 mem # 0: /home/oraprod/PROD/data/log01a.dbf

  Currently journal # 1, seq # 3387 mem # 1: /home/oraprod/PROD/data/log01b.dbf

  Beginning log switch checkpoint up to RBA [0xd3c.2.10], RCS: 5978324634623

  Thread 1 Advanced for you connect to sequence 3388 (switch LGWR)

  Currently Journal # 2 seq # 3388 mem # 0: /home/oraprod/PROD/data/log02a.dbf

  Currently Journal # 2 seq # 3388 mem # 1: /home/oraprod/PROD/data/log02b.dbf

  Mon Jan 25 15:11:01 2016

  LNS: Standby redo log file selected for thread 1 sequence 3388 for destination LOG_ARCHIVE_DEST_2

  Mon Jan 25 15:11:04 2016

  Archived journal 6791 extra for each sequence 1 3387 ID thread entry 0 x 12809081 dest 1:

  Mon Jan 25 15:11:17 2016

  Filled checkpoint up to RBA [0xd3c.2.10], RCS: 5978324634623

  Mon Jan 25 15:13 2016

  Additional checkpoint up to RBA to RBA [0xd3c.1a8e82.0] [0xd3c.18d210.0], current journal line

  Mon Jan 25 15:13:04 2016

  Thread 1 cannot allot of new newspapers, sequence 3389

  Private stream flush is not complete

  Currently Journal # 2 seq # 3388 mem # 0: /home/oraprod/PROD/data/log02a.dbf

  Currently Journal # 2 seq # 3388 mem # 1: /home/oraprod/PROD/data/log02b.dbf

  Beginning log switch checkpoint up to RBA [0xd3d.2.10], RCS: 5978324673444

  Thread 1 Advanced for you connect to 3389 (switch LGWR) sequence

  Currently journal # 1, seq # 3389 mem # 0: /home/oraprod/PROD/data/log01a.dbf

  Currently journal # 1, seq # 3389 mem # 1: /home/oraprod/PROD/data/log01b.dbf

  Mon Jan 25 15:13:07 2016

  LNS: Standby redo log file selected for thread 1 sequence 3389 for destination LOG_ARCHIVE_DEST_2

  Mon Jan 25 15:13:09 2016

  Archived journal 6793 extra for each sequence 1 3388 ID thread entry 0 x 12809081 dest 1:

  Mon Jan 25 15:13:11 2016

  Filled checkpoint up to RBA [0xd3d.2.10], RCS: 5978324673444

  Is it a sign of botteneck? Users complained that each 15:00 they encounter performance degradation.

  Kind regards

  JC

  Jenna_C wrote:

  We have two groups of size 1 GB log and our archiving logs generates 13 newspapers every hour.

  Is it a sign of botteneck? Users complained that each 15:00 they encounter performance degradation.

  If your users are complaining about slow at 15:00 then look at what is happening on the system at this point in time.  Use the Active Session history, providing that you have the license, at 3 pm to see if there is a problem.  Or take snapshots of the AWR manually around 15:00 in an attempt to capture any anomaly.  Or capture instant V$ SESSION yourself manually around 15:00 and see so many sessions is pending or not.

  I have seen problems like this on shared physical infrastructure systems where the activity on a system can interfere with each other.  Attention to this kind of thing, because if it is caused by another system then you will never find the cause of the problem inside the Oracle.  Examples include backups are made at a specific time, causing large amounts of sequential disk i/o and slows down access to the drive for everything else.  Such a scenario could cause some writing redo log time to increase significantly.  Or the backup could be another system completely who shares the same disks in the same drive bay or the SAN and Bay drives or SAN is inundated by applications e/s backup.  But as it's another system, that you will not see what is happening on your Oracle database system.

  Or a virtualized with environment sharing CPU and memory where another virtual machine running a CPU intensive job and steals the Oracle system CPU capacity.  I saw the same thing with memory so - Vmware has a strange way to virtualize memory, which means it can use all of the memory on the physical hardware and run out, and then he had to intervene to release somehow little memory.  What ends up slowing things down.

  It could be a network problem - is something that the network of the floods at 15:00?  Still, backups can be a common cause of this, during the transfer of data over the network to another system.  Is there a work extracted regularly data which runs at 15:00 for example extract given OLTP to feed into a data warehouse, which is copied on the network?

  It could be many different things, causing the slowdown.  I would definitely recommend watching any expected sessions know and see if this gets worse around 15:00, or if it remains the same, in which case the problem is elsewhere.  You must also eliminate everything else - the disks, network, etc.

  Good luck

  John Brady

• Will be online Redo Logs to sleep never used?

  Version: 11 GR 2

  OS: Oracle Linux 6.5

  Pending a physical DB, newspapers online redo will be ever written because the author of newspaper on the standby site will be inactive. Right?

  In the DB in waiting, only forward Redo logs will be written to by RFS. Right?

  Pending a physical DB, newspapers online redo will be ever written because the author of newspaper on the standby site will be inactive. Right?

  ORL will be used when the database starts to behave like a primary database and LGWR writes the data to the newspaper for the ORL buffer again.

  They are not used while the database behaves like a physical standby.

  In the DB in waiting, only forward Redo logs will be written to by RFS. Right?

  Yes, the process RFS writes that data to the standby redo redo logs on the standby database.

  Just for your reference https://docs.oracle.com/cd/B19306_01/server.102/b14239/log_apply.htm#i1022881

  HTH

  -Jonathan Rolland

• in oracle redo-log, QMD means fast Multi delete, but what do I do to generate such redo log?

  to QMI(Quick multi-insert), I can run the following SQL to generate the QMI redo log:

  Insert into t1 select * from t2;

  or use forall sentence to achieve this goal.

  but I do not know how to generate log redo QMD, can someone help me?

  Hello

  When I restore after an insert... Select I get the following type of repeat folder:

  PROGRESS RECORD roll - Threading: 1 RBA: LEN 0x000e78.00005fb5.00cc: 0x00e8 VLD: 0x01 CON_UID: 0

  RCS: 0x0000.0b8cc83a SUBSCN: 1 08/12/2015 23:07:08

  CHANGE CON_ID:0 #1 TYP:0 CLS:1 AFN: 6 x 0000 .0b8cc837 SEQ:4 OP:11.12 ENC:0 SCN:0 OBJ:116277 DBA:0x018264c2 RBL: 0 FLG:0 X 0000

  KTB Redo

  OP: worm 0 x 02: 0 x 01

  compat bit: 4 (post-11) stuffing: 1

  OP: C uba: 0x01001a86.0d68.0c

  Op KDO code: line QMD disabled dependencies

  xType: XR flags: 0 x 00000000 bdba: hdba 0x018264c2: 0 x 01825972

  itli: 1 ispac: 0 maxfr: 4858

  Tabn: 0 lock: nrow equals 0: 16

  location [0]: 54

  location [1]: 55

  slot [2]: 56

  slot [3]: 57

  Housing [4]: 58

  Housing [5]: 59

  Housing [6]: 60

  slot [7]: 61

  slot [8]: 62

  Housing [9]: 63

  slot [10]: 64

  slot [11]: 65

  Housing [12]: 66

  slot [13]: 67

  slot [14]: 68

  slot [15]: 69

  CHANGE CON_ID:0 #2 TYP:0 CLS:24 AFN: 4 x 0000 .0b8cc83a SEQ:13 OP:5.6 ENC:0 SCN:0 OBJ:4294967295 DBA:0x01001a86 RBL: 0 FLG:0 X 0000

  ktubu again: am: rci 0: 12 opc: 11.1 objn: 116277 objd: tsn 116277: 4

  Cancel type: Regular cancel cancel type: user cancellation made last buffer split: No.

  Undo tablespace: No.

  0x00000000

  ktuxvoff: 0x17cc ktuxvflg: 0x0002

  Kind regards

  Franck.

  PS: this is the replication product that I think?

• Best location for the archived redo logs

  Hello

  I am OOF instructions and I want to make life for the DBA, that looks like my job easier.

  So, as the title says what is the location of the "standard/Best Practice' for archived redo logs? particularly the dest_1 which is usually local on the same server

  Thank you.

  Hello

  For you, I recommend the use of the Flash/fast recovery area.

  Configuration of the archived of Redo Log locations

  Oracle recommends that you use fast area recovery to an archive location, because the archived logs are managed automatically by the database. The file names generated for the newspapers archived in the fast recovery area correspond to Oracle managed files and are not determined by the parameter LOG_ARCHIVE_FORMAT . Whatever archiving scheme you choose, it is always advisable to create multiple copies of archived redo logs.

  Ref 1:

  http://docs.Oracle.com/CD/E11882_01/backup.112/e10642/rcmconfb.htm#CHDEHHDH

  Ref 2:

  http://docs.Oracle.com/CD/E11882_01/server.112/e17157/unplanned.htm#BABEEEFH

  REF 3:

  http://docs.Oracle.com/CD/E11882_01/backup.112/e10642/rcmconfb.htm#BRADV89418

  Kind regards

  Juan M