Data Guard Standby Lost A Drive Or Two

So a data guard standby is somewhat disgruntled because the SAN imploded and the standby lost a couple mount points. It wasn't running ASM so no self-repairing can occur. Since this is the first time I've had a serious issue with a standby and since much of what I do is restricted (RMAN complains because it's a standby and read-only)... I'm just curious if someone could point me towards a comprehensive "what to do when your STANDBY has issues" book, document, or link. They are surprisingly scarce. Most comments online simply say (paraphrased) "scorch the earth, salt the fields, and rebuild fresh."
Thank you

Hello;
Depends what you lost.
1. You can just rebuild.
2. You might be able to use RMAN roll forward.
http://www.oracle-ckpt.com/?s=standby+scn&op.x=30&op.y=17
http://emrebaransel.blogspot.com/2008/12/recover-dataguard-from-lost-archivelog.html
3. If you are Oracle 11 you can use RMAN active duplicate to rebuild.
http://www.visi.com/~mseberg/standby_creation_from_active_database_using_rman.html
4. Other
Recover Lost Datafile On Standby
http://hemendraoracleblogs.blogspot.com/2012/11/recover-lost-datafile-from-standby.html
http://allthingsoracle.com/data-guard-physical-standby-database-best-practices-part-ii/
http://stackoverflow.com/questions/8135393/how-do-you-re-duplicate-a-broken-physical-standby-database
Best Regards
mseberg

Similar Messages

Data guard standby restart in rocover mode

Hi,
I inherited a data guard database and found that the secondary DB has been out of sync with primary. I perform SCN recovery to the secondary and now they are fine syncing. I figured I had to manually put the standby in recover mode after open in read only mode in the one I worked on which is our qual environment where as it is not like that in the prod. Can anyone advise what I need to do to have this standby in recovery mode even after the DB go down for maintenance and come back on?
Thanks for your input,
fakord

Hi,
If you are using database guard broker then it will automatically starts the apply as it controls the apply. If not probably a shell script to start the standby database followed by this line can help.
alter database recover managed standby database using current logfile disconnect from session;
HTH,
Pradeep

Data Guard: standby archivelog files not the same as the primary archivelog

From reading the Data Guard doco, one gets the impression that that the standby redo archive logs are just a mirror of those from the primary, that the shipped and applied archived redo logs from the primary are what end up in the standby archive directories.
However I have noted that the files are not the same (md5sum is different), which is concerning as using Oracle's archive log naming recommendation, the files have identical names.
Does this mean that the Standbys redologs can not be used as a source of redo logs if one had to restore using a backup of the primary database?
i.e. Can you recover using an RMAN backup of the primary, and redo logs from the standby?

Hi MSK, I agree, that's exactly what I thought. I'm confused though why the logs do not match. Perhaps it is just metadata in the log that has changed, not redo data.
Anyone have any knowledge on why the archivelogs are different on the standby to those on the primary?

Active data guard standby database

Hello,
we have to choose which technology to use for maintain Oracle backup center synchronized with main center. This backup center should be read only, and data generated in main should be replicated into backup. So I would like to use Active Data Guard, but I have a possible issue with reading in backup center, because reading is not just select, but filling(inserting data) some temporary tables (from client side, and from procedures in database). Will Standby database in backup center allow this if Active Data Guard is used. We would go with Oracle 11gR2.
Also other suggestions on replication technologies (I know there are RAC, ADG, Oracle Streams, Golden Gate), or any document which would help us decide which of these technologies would suit our needs the best.
Thank You.

Hi,
user12121832 wrote:
we have to choose which technology to use for maintain Oracle backup center synchronized with main center. This backup center should be read only, and data generated in main should be replicated into backup. So I would like to use Active Data Guard, but I have a possible issue with reading in backup center, because reading is not just select, but filling(inserting data) some temporary tables (from client side, and from procedures in database). Will Standby database in backup center allow this if Active Data Guard is used. We would go with Oracle 11gR2. You can use Standby Database to make yours Backup.
In 11.1 or later you can use feature "Snapshot Standby Databases" if this write on database is temporary (i.e you will not use the new data) and at backup time you not need use Snapshot Standby Database feature.
A snapshot standby database is a physical standby database that you temporarily convert into an updatable standby database. You can use snapshot standby databases as clones or test databases to validate new functionality and new releases, and when finished you then convert the database back into a physical standby. While running in the snapshot standby database role, it continues to receive and queue redo data so that data protection and the RPO are not sacrificed.
A snapshot standby database has the following characteristics:
* A snapshot standby database cannot be the target of a switchover or failover. A snapshot standby database must first be converted back into a physical standby database before performing a role transition to it.
* A snapshot standby database cannot be the only standby database in a Maximum Protection Data Guard configuration.
Note:
Flashback Database is used to convert a snapshot standby database back into a physical standby database. Any operation that cannot be reversed using Flashback Database technology will prevent a snapshot standby from being converted back to a physical standby.
http://docs.oracle.com/cd/E11882_01/server.112/e17157/unplanned.htm#HAOVW11832
Also other suggestions on replication technologies (I know there are RAC, ADG, Oracle Streams, Golden Gate), or any document which would help us decide which of these technologies would suit our needs the best.If you need "standby" database opened in read-write all time the best option is to use Golden Gate.
Regards,
Levi Pereira

Data Guard & Standby difference

Hi All,
Can any one tell me what is difference between the DG & standby, is any document avaliable for the reference.
I have worked on standby but not on DG.
Thanks,
Abhay.

Hi,
Data guard is nothing but a marketing term for standby resulting from an extra interface to manage the standby database. Now Physical Standby on the other hand is a type of the standby database within a dataguard configuration. If you mean the dataguard broker, then it is an added feature of Data Guard.
DBAs talks for standby database, while marketing fellows talks in term of "Data Guard"
Regards
Girish Sharma

Data Guard Standby DB Duplicate entries in v$archived_log - apply stopped

Hi
I've got a perplexing problem... I have a number of archive logs that seem to have gotten duplicated (there are no other log files past or prior to this particular point). This duplication however has stopped the apply of the logs to the standby database. (Getting the 'apply' to continue is of paramount importance)
Any suggestions on how to remove the 'duplicate' entries on the standby database would be appreciated.
Here is the output from my query on the v$archived_log view:
1130 20100601 11:02:38 20100601 11:53:05 NO YES YES D
1131 20100601 11:53:05 20100601 12:21:04 NO YES YES D
1132 20100601 12:21:04 20100601 12:44:33 NO YES YES D
1133 20100601 12:44:33 20100601 13:08:52 NO YES NO A <--- Duplication starts
1133 20100601 12:44:33 20100601 13:08:52 NO YES NO A
1133 20100601 12:44:33 20100601 13:08:52 NO YES NO A
1133 20100601 12:44:33 20100601 13:08:52 NO YES NO A
1133 20100601 12:44:33 20100601 13:08:52 NO YES NO A
1133 20100601 12:44:33 20100601 13:08:52 NO YES NO A
1134 20100601 13:08:52 20100601 13:33:32 NO YES NO A
1134 20100601 13:08:52 20100601 13:33:32 NO YES NO A
SEQUENCE# First|Time Next|Time STA ARC APP S
1134 20100601 13:08:52 20100601 13:33:32 NO YES NO A
1135 20100601 13:33:32 20100601 13:52:01 NO YES NO A
1135 20100601 13:33:32 20100601 13:52:01 NO YES NO A
1135 20100601 13:33:32 20100601 13:52:01 NO YES NO A
1136 20100601 13:52:01 20100601 14:10:55 NO YES NO A
1136 20100601 13:52:01 20100601 14:10:55 NO YES NO A
1136 20100601 13:52:01 20100601 14:10:55 NO YES NO A
1137 20100601 14:10:55 20100601 14:58:02 NO YES NO A
1137 20100601 14:10:55 20100601 14:58:02 NO YES NO A
1137 20100601 14:10:55 20100601 14:58:02 NO YES NO A
1138 20100601 14:58:02 20100601 15:28:56 NO YES NO A
SEQUENCE# First|Time Next|Time STA ARC APP S
1138 20100601 14:58:02 20100601 15:28:56 NO YES NO A
1138 20100601 14:58:02 20100601 15:28:56 NO YES NO A
1139 20100601 15:28:56 20100601 16:22:01 NO YES NO A
1139 20100601 15:28:56 20100601 16:22:01 NO YES NO A
1139 20100601 15:28:56 20100601 16:22:01 NO YES NO A
1140 20100601 16:22:01 20100601 18:08:30 NO YES NO A
1140 20100601 16:22:01 20100601 18:08:30 NO YES NO A
1140 20100601 16:22:01 20100601 18:08:30 NO YES NO A
1141 20100601 18:08:30 20100601 18:08:31 NO YES NO A
1141 20100601 18:08:30 20100601 18:08:31 NO YES NO A
1141 20100601 18:08:30 20100601 18:08:31 NO YES NO A
SEQUENCE# First|Time Next|Time STA ARC APP S
1141 20100601 18:08:30 20100601 18:08:31 NO YES NO A
1141 20100601 18:08:30 20100601 18:08:31 NO YES NO A
1142 20100601 18:08:31 20100601 22:00:14 NO YES NO A
1142 20100601 18:08:31 20100601 22:00:14 NO YES NO A
1142 20100601 18:08:31 20100601 22:00:14 NO YES NO A
1143 20100601 22:00:14 20100601 22:15:24 NO YES NO A
1143 20100601 22:00:14 20100601 22:15:24 NO YES NO A
1143 20100601 22:00:14 20100601 22:15:24 NO YES NO A
1144 20100601 22:15:24 20100602 02:57:01 NO YES NO A
1144 20100601 22:15:24 20100602 02:57:01 NO YES NO A
1144 20100601 22:15:24 20100602 02:57:01 NO YES NO A
SEQUENCE# First|Time Next|Time STA ARC APP S
1145 20100602 02:57:01 20100602 03:33:21 NO YES NO A
1145 20100602 02:57:01 20100602 03:33:21 NO YES NO A
1145 20100602 02:57:01 20100602 03:33:21 NO YES NO A
1146 20100602 03:33:21 20100602 07:38:50 NO YES NO A
1146 20100602 03:33:21 20100602 07:38:50 NO YES NO A
1146 20100602 03:33:21 20100602 07:38:50 NO YES NO A
1147 20100602 07:38:50 20100602 08:23:19 NO YES NO A <-- Duplication ends....
1148 20100602 08:23:19 20100602 08:57:02 NO YES NO A
1149 20100602 08:57:02 20100602 09:41:02 NO YES NO A
1150 20100602 09:41:02 20100602 10:30:40 NO YES NO A
1151 20100602 10:30:40 20100602 10:57:11 NO YES NO A
Shipping of logs are continuing normally (we had to stop McAfee on the machines -- wrecked havoc with the replicaton -- not sure if the duplication can be attributed to the virus checker too)
Thanks for the attention to this problem...
Kind Regards
Andries

Thanks for posting - I've not made any changes to this system. The archive log listing I posted is from my standby database and after it got itself out of whatever not it got itself into in the first place, it applied all the logs... even these duplications have updated to YES (in the applied column). Very strange occurence... have not seen it since this. Hope it's a once off aberation.
since this is no longer a problem... I'll mark this question as answered.
Thanks to all who looked at it.

Data Guard : Standby Redo Log CONFUSION

Trying to set up test Standby db on 10.2.0
I am well confused about below step 3.1.3, how is the normal redo linked with standby redo, should standby not be members of orginal redo groups?
http://download.oracle.com/docs/cd/B19306_01/server.102/b14239/create_ps.htm#i1225703
Original redo logs:
SQL> select * from v$log;
    GROUP#    THREAD# SEQUENCE#      BYTES    MEMBERS ARC STATUS           FIRST_CHANGE# FIRST_TIME
         1          1         28   52428800          1 YES INACTIVE                375136 22-NOV-07
         2          1         29   52428800          1 YES INACTIVE                375138 22-NOV-07
         3          1         30   52428800          1 NO CURRENT                 375143 22-NOV-07I added below from notes:
ALTER DATABASE ADD STANDBY LOGFILE GROUP 10
('/u01/oracle/oradata/db01/redo01_stb.log') SIZE 50M;
ALTER DATABASE ADD STANDBY LOGFILE GROUP 11
('/u02/oracle/oradata/db01/redo02_stb.log') SIZE 50M;
ALTER DATABASE ADD STANDBY LOGFILE GROUP 12
('/u03/oracle/oradata/db01/redo03_stb.log') SIZE 50M;After few alter system switch logfile; I still have:
    GROUP#    THREAD# SEQUENCE# ARC STATUS
        10          0          0 YES UNASSIGNED
        11          0          0 YES UNASSIGNED
        12          0          0 YES UNASSIGNEDAll are UNASSIGNED, should one standby group not be ACTIVE like the above link shows.
Many think for any help

First things first:
From the Docs.:
"Minimally, the configuration should have one more standby redo log file group than the number of online redo log file groups on the primary database. However, the recommended number of standby redo log file groups is dependent on the number of threads on the primary database. Use the following equation to determine an appropriate number of standby redo log file groups:
(maximum number of logfiles for each thread + 1) * maximum number of threads
Using this equation reduces the likelihood that the primary instance's log writer (LGWR) process will be blocked because a standby redo log file cannot be allocated on the standby database. For example, if the primary database has 2 log files for each thread and 2 threads, then 6 standby redo log file groups are needed on the standby database."
You are 1 short!

Oracle 10gR2 Data guard physical or logical standby server?

Hi
We are planing to implement an Oracle 10gR2 data guard standby server for DR purposes, I found out that there are two type of standby server which is logical and physical standby server. I want to know which one is preferable? in term of complexity of setup and maintenance?
regards

Well it depends on what you mean by maintenance. I found the physical standby to be very little trouble at all ; however the logical standby has restrictions on it that the physical standby does not. In essence the physical standby merely digest archive logs; where as the logical standby uses logminer like functionality to process sql statements much like Oracle streams.
Hope that helps,
-JR jr.

Rman clone ? duplicate database ? data guard ? standby database?

hey there !!
What is the difference between Rman clone , duplicate database , data Guard , standby db?
1 more thing , standby database and duplicate database are same??

RMAN Clone : you take backup of a database on one server ,then take it to other server ,again use RMAN to restore the database,A lot manual intervention is needed in this method
RMAN DUPLICATE: same as above just that manual intervention is less. With 11gR2 there active database duplication in which you can clone your test database directly from production database without taking and then transferring the backup to test server.
Standby Database :they are of two types Physical and logical.The setup is done using RMAN with any of the above methods.Standby is the in-sync (can have some delay) version of primary database either block by block which is physical standby configuration or SQL statment wise which is logical standby.
Data Guard:This is also a standby configuration.Here the failover and switchover are automated with the help of DGMGRL utility.In manual standby this same can be done via SQL*PLUS using sql commands

Difference between physical standby database and Data Guard

Hi guys,
Today I attended one interview. They asked me a difference between the Oracle Physical Standby database and Oracle Dataguard?
Upto my knowledge Oracle Physical Standby Database is one type of Standby Database can be configured using Oracle Databaguard Feature.
Can you people calrify me on this.
Regards,
Vijayaraghavan K

hello,
yes you are right upto some extent...here is para form oracle doc
Oracle Data Guard is the most effective and comprehensive
data availability, data protection, and disaster recovery solution
for enterprise databases. It provides the management, monitoring,
and automation software infrastructure to create and maintain one
or more synchronized standby databases to protect data from failures,
disasters, errors, and corruptions.
Data Guard standby databases can be located at remote disaster
recovery sites thousands of miles away from the production data center,
or they may be located in the same city, same campus,
or even in the same building. If the production database
becomes unavailable because of a planned or an unplanned outage,
Data Guard can switch any standby database to the production role,
thus minimizing downtime and preventing any data loss.oracle data guard is maximum availiblity solution which maintain physical standby databases and in case of failure of production can switch production role to and standby database.
read it here
http://www.oracle.com/technology/deploy/availability/htdocs/DataGuardOverview.html
thanks and regards
VD
Edited by: Dixit on Aug 31, 2009 1:35 AM
Edited by: Dixit on Aug 31, 2009 1:36 AM

Data Guard - dual configuration

Hi,
we are running Data Guard configuration (Windows Server 2008 R2) with physical standby database in place, using FSFO configuration with DG Broker - observer process on separate machine (10.2.0.5). My question is, if it's safe and possible to create another, separate Data Guard configuration (11.2.0.3) on the same machines (2x database server - databases, 1x client PC - observer)?
Thank you,
t.

Hello;
I've done it. Be it with Linux but that should not matter. I had two Oracle homes, one for each version and a script to set as needed. I ran into this during database upgrades too. Some of the databases were ready and some were not, so I was running 11.2.0.2 and 11.2.0.3 on the same server at the same time, both with Data Guard. No issues. The main thing was to use one listener and include the entries for the lower version in the file. I always start and stop the listener from the higher version.
A few years ago I still had some Oracle 10 databases and they worked just fine on a server with Oracle 11. The separate Oracle homes is the key. So on Linux I had no issue running Data Guard on the same server with two versions of Oracle.
Best Regards
mseberg
This is older, but worth a look : ( Using Multiple Oracle Homes )
http://docs.oracle.com/cd/B10500_01/em.920/a96697/moh.htm
Edited by: mseberg on May 31, 2013 7:49 AM

DB link problem between active Data Guard and report application database

My database version in 11.2.0.2.0 and OS is Oracle Solaris 10 9/10.
I am facing a problem in my Active data guard Database for reporting purpose. Active Data guard information is as below.
SQL> select name, database_role, open_mode from v$database;
NAME DATABASE_ROLE OPEN_MODE
ORCL PHYSICAL STANDBY READ ONLY WITH APPLY
Problem detail is below
I have created a db link (Name: DATADB_LINK) between active data guard and report application database for reporting purpose.
SQL> create database link DATADB_LINK connect to HR identified by hr using 'DRFUNPD';
Database link created.
But when I run a query using db link from my report application database I got this below error.
ORA-01555: snapshot too old: rollback segment number 10 with name "_SYSSMU10_4261549777$" too small
ORA-02063: preceding line from DATADB_LINK
Then I check Active Data Guard database alart log file and get below error
ORA-01555 caused by SQL statement below (SQL ID: 11yj3pucjguc8, Query Duration=1 sec, SCN: 0x0000.07c708c3):SELECT "A2"."BUSINESS_TRANSACTION_REFERENCE","A2"."BUSINESS_TRANSACTION_CODE",MAX(CASE "A1"."TRANS_DATA_KEY" WHEN 'feature' THEN "A1"."TRANS_DATA_VALUE" END ),MAX(CASE "A1"."TRANS_DATA_KEY" WHEN 'otherFeature' THEN "A1"."TRANS_DATA_VALUE" END )
But the interesting point if I run the report query directly in Active Data Guard database, I never got error.
So is it a problem of DB link between active Data Guard and other database?

Fazlul Kabir Mahfuz wrote:
My database version in 11.2.0.2.0 and OS is Oracle Solaris 10 9/10.
I am facing a problem in my Active data guard Database for reporting purpose. Active Data guard information is as below.
SQL> select name, database_role, open_mode from v$database;
NAME DATABASE_ROLE OPEN_MODE
ORCL PHYSICAL STANDBY READ ONLY WITH APPLY
Problem detail is below
I have created a db link (Name: DATADB_LINK) between active data guard and report application database for reporting purpose.
SQL> create database link DATADB_LINK connect to HR identified by hr using 'DRFUNPD';
Database link created.
But when I run a query using db link from my report application database I got this below error.
ORA-01555: snapshot too old: rollback segment number 10 with name "_SYSSMU10_4261549777$" too small
ORA-02063: preceding line from DATADB_LINK
Then I check Active Data Guard database alart log file and get below error
ORA-01555 caused by SQL statement below (SQL ID: 11yj3pucjguc8, Query Duration=1 sec, SCN: 0x0000.07c708c3):SELECT "A2"."BUSINESS_TRANSACTION_REFERENCE","A2"."BUSINESS_TRANSACTION_CODE",MAX(CASE "A1"."TRANS_DATA_KEY" WHEN 'feature' THEN "A1"."TRANS_DATA_VALUE" END ),MAX(CASE "A1"."TRANS_DATA_KEY" WHEN 'otherFeature' THEN "A1"."TRANS_DATA_VALUE" END )
But the interesting point if I run the report query directly in Active Data Guard database, I never got error.
So is it a problem of DB link between active Data Guard and other database?
Check this note which is applicable for your environment
*ORA-01555 on Active Data Guard Standby Database [ID 1273808.1]*
also
http://asktom.oracle.com/pls/asktom/f?p=100:11:0::::P11_QUESTION_ID:8908307196113

Data Guard - Different Platforms

Hello,
I'm currently looking at Oracle's Data Guard/Standby DB option. (Standard Edition Version) Our primary database is on a Sun Solaris Sparc system and we want to have our Standby on an Sun Solaris Intel system. Is this possible with Sun Sparc? Will I run into issues attempting to apply the redo logs?
Help??? :)
Edited by: user10448714 on Aug 13, 2009 9:47 AM

Check these Metatlink notes they may help you out here.
execute this SQL for a database on the 2 platforms
SQL> select platform_id, platform_name from v$database;
You will need to check the platform_id for solaris it is 2, if the intel solaris is 2 then according to the document it should work.
For a complete list of supported platform/OS combinations see Metalink Note 413484.1
See Oracle Metalink Notes 395982.1 and/or 414043.1 for more exact restrictions and further details.

Difference between oracle Data Guard and Replication

Hello Friends,
I would like to know the main difference between oracle replication and data guard. Theoretically both are same then where is the difference. Can replication or dataguard be configured in a single PC.
Thanks

Dataguard :
Oracle Data Guard is primarily for high availability, data protection, and disaster recovery. An Oracle Data Guard Standby instance provides transactionally consistent copies of the production database, in the event of an unplanned outage or instance maintenance the Data Guard Primary Role Standby Role can be switched so the Standby instance can provide Production services.
Replication :
Oracle Advanced Replication and Oracle Streams are Oracle's solution for replicating data and internal structures to remote databases. Advanced Replication is primarily trigger based where Streams is a rule/handler based replication process and considerably more efficient and less error prone than Advanced Replication
refer,
http://www.xtivia.com/database-management/replication/oracle-replication
Thanks

Using data guard for migration pruposes.

Hi we are looking to consolidate a couple of databases onto a single server. One of the databases is our bi database which is currently running on a single server and it is running oracle 10.1.0.3 on suse linux 64 bit.
We also have an oracle 2 node cluster running on RHEL4 64 bit linux itanium. which is our sote database.
We are looking to create a physical standby server of the store database onto a new server and storage array and also we are looking to migrate the bi database over to the same server/storage platform and would like to use ASM as the storage option for both of the database once they are migrated and also replicated to the new server.
I guess my question is...can I run multiple different versions of oracle on the same server or is this just a bad idea? we know that we need to have the same version 10.2.0.3 for the physical standby server which is a different version than the bi datatabse so would it be better idea to upgrade the bi database to also be running 10.2.0.3 before we are to try and replicate it to the new hardware?
I hope this is not confusing,
Thx.

Lets start with what physical Data Guard standby is ... it is the creation of a row-for-row, block-for-block duplicate of a primary production database.
If the version are not identical that is impossible.
If the data contained within them is not identical that is impossible.
If you want to merge databases I would suggest Transportable Tablespaces ... nothing is faster and safer.
Once you have you data merged you can then create a physical Data Guard standby for purposes of replication if you wish.

Data Guard Standby Lost A Drive Or Two

Similar Messages

Maybe you are looking for