Re: Oracle Data Guard
Comments in-line
At 00:49 11-12-03 -0800, you wrote:
But the last time I looked at it, you had
to enable supplemental logging at the
database level if you wanted to use logical
standby. Two side effects -
Yes, indeed, supplemental logging must be switched on.
1) As you said, you need a unique/primary key,
and database supplemental logging copies
such a key into the redo for every change to a row:
but if there is not uk/pk, then the whole row is copied.
This will, of course, give you significant more redo when rows are long,
and UK/{K not there.
So, try to find some unique identification, and, if possible, consider
creating a meaningless PK.
Might be cheaper than having complete rows copied into redo.
2) The copy into redo is engineered by copying into
the UNDO first - and since changes to GTTs are
recorded into the UNDO, this means you get an
extra volume of UNDO, hence REDO on all changes
to 'supposed to be low-cost' GTTs.
Sorry Jonathan, might be a stupid question: what are GTT's?
The other feature of logical standby is that Oracle
scrapes the redo log to generate LCRs (logical change
records) which are then checked against your 'Streams
rule-sets' - and then written into the local database
for propagation to the remote via AQ mechanisms.
Yes, but this is performed at the standby database, not at the primary.
IMHO this is the major difference between streams and Data Guard / Logical
Standby: With streams the LCR is genereated locally and queued to the
destinations, with Data Guard the redo gets copied to the standby and the
redo scraping/LCR generations is done overthere. So, apart from all your
findings about generating extra (possibly/probably significant) UNDO and
REDO, the primary is only hurt by sending the REDO's to the Standby(s).
The overheads could be quite significant.
Yes!. So test your individual situation before going to production. It's
all a trade-off. Streams, Advanced Replication, Data Guard Physical Standby
with daily refreshes, Data Guard Logical Standby with extra overhead, they
all com with your Oracle Licences. Probably even Shareplex could help you
out, but will cost you plenty of money. There is no panacee, you have to
find the solution that fits your needs AND budget. If your budget is too
low, you have to adapt your needs, or quality will suffer.
Regards, Carel-Jan Engel
-- There will allways be another 10 last bugs --
Regards
Jonathan Lewis
http://www.jlcomp.demon.co.uk
The educated person is not the person
who can answer the questions, but the
person who can question the answers -- T. Schick Jr
One-day tutorials:
http://www.jlcomp.demon.co.uk/tutorial.html
Three-day seminar:
see http://www.jlcomp.demon.co.uk/seminar.html
UK___November
The Co-operative Oracle Users' FAQ
http://www.jlcomp.demon.co.uk/faq/ind_faq.html
- Original Message -
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
Sent: Tuesday, December 09, 2003 9:59 PM
Hi Vi,
Rows NEED unique identification. So, if there are bunches of raw data with
no unique identifier whatsoever (remember, rowid is not allowed) LSB can't
generate a where-clause what row to update or delete on the SB database.
It's generating SQL based on redolog info, and has to come up with an
UPDATE table SET WHERE unique id = unique id. The unique id may
be a multi-column key. There is an escape. Enabling supplemental logging
can add extra info to do the unique identification, when no usable keys are
available. This will cause some extra logging to be generated, there ain't
no such thing as a free lunch. For detailed information read chapter 4.1.5
4.1.6 in the Oracle Data Guard Concepts and Administration manual, part
no. A96653-02.
regards, Carel-Jan
At 15:54 8-12-03 -0800, you wrote:
--
Please see the official ORACLE-L FAQ: http://www.orafaq.net
--
Author: Jonathan Lewis
INET: [EMAIL PROTECTED]
Fat City Network Services-- 858-538-5051 http://www.fatcity.com
San Diego, California-- Mailing list and web hosting services
-
To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (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.net
--
Author: Carel-Jan Engel
INET: [EMAIL PROTECTED]
Fat City Network Services-- 858-538-5051 http://www.fatcity.com
San Diego, California-- Mailing list and web hosting services
-
To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (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
Re: Oracle Data Guard
Can you clarify a couple of points for me. The SDU (session data unit) is presumably the packet size that the Oracle client and server want to pass back and forth - which is presumably the maximum size the one synchronous dialogue unit will be. The TDU (transport data unit) is presumably the predicted size of the transport maximum unit of data transfer (MTU). a) Why does Oracle need to know anything about the underlying transport mechanism ? b) If I set the SDU to the largest legal value (possibly 32K, perhaps 64K) the server task switch will occur after building and sending that packet - is there any good reason why I shouldn't do that. After all, if the transport simply accepts the 64K packet and gets it to the other end of the wire (not yet to the client session, just to the receiving transport layer) as rapidly as possible does it matter to Oracle whether the transport is using 1.5K or 8K packets. The fact that the transport layer doesn't have to work its packet synchronously means that some overheads have disappeared as far as Oracle is concerned. Regards Jonathan Lewis http://www.jlcomp.demon.co.uk The educated person is not the person who can answer the questions, but the person who can question the answers -- T. Schick Jr One-day tutorials: http://www.jlcomp.demon.co.uk/tutorial.html Three-day seminar: see http://www.jlcomp.demon.co.uk/seminar.html UK___November The Co-operative Oracle Users' FAQ http://www.jlcomp.demon.co.uk/faq/ind_faq.html - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Wednesday, December 10, 2003 3:44 PM Hi, Guang, Look up SDU and TDU in Oracle documentation Network configuration. You set them in tnsnames.ora and listener.ora, not sqlnet.ora. protocol.ora allows you to modify some procotol-specific parameters. In addition, in your client application, you can choose a sensible array fetch size, such as arraysize in sqlplus (in fact, sqlplus arraysize changes more than just network data chunk size). You can't magically increase the network transfer rate by lowering network latency. But you can indirectly increase the rate by other means, such as buffering slightly more data in one chunk. Yong Huang -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Jonathan Lewis INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
Re: Oracle Data Guard
But the last time I looked at it, you had to enable supplemental logging at the database level if you wanted to use logical standby. Two side effects - 1) As you said, you need a unique/primary key, and database supplemental logging copies such a key into the redo for every change to a row: but if there is not uk/pk, then the whole row is copied. 2) The copy into redo is engineered by copying into the UNDO first - and since changes to GTTs are recorded into the UNDO, this means you get an extra volume of UNDO, hence REDO on all changes to 'supposed to be low-cost' GTTs. The other feature of logical standby is that Oracle scrapes the redo log to generate LCRs (logical change records) which are then checked against your 'Streams rule-sets' - and then written into the local database for propagation to the remote via AQ mechanisms. The overheads could be quite significant. Regards Jonathan Lewis http://www.jlcomp.demon.co.uk The educated person is not the person who can answer the questions, but the person who can question the answers -- T. Schick Jr One-day tutorials: http://www.jlcomp.demon.co.uk/tutorial.html Three-day seminar: see http://www.jlcomp.demon.co.uk/seminar.html UK___November The Co-operative Oracle Users' FAQ http://www.jlcomp.demon.co.uk/faq/ind_faq.html - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Tuesday, December 09, 2003 9:59 PM Hi Vi, Rows NEED unique identification. So, if there are bunches of raw data with no unique identifier whatsoever (remember, rowid is not allowed) LSB can't generate a where-clause what row to update or delete on the SB database. It's generating SQL based on redolog info, and has to come up with an UPDATE table SET WHERE unique id = unique id. The unique id may be a multi-column key. There is an escape. Enabling supplemental logging can add extra info to do the unique identification, when no usable keys are available. This will cause some extra logging to be generated, there ain't no such thing as a free lunch. For detailed information read chapter 4.1.5 4.1.6 in the Oracle Data Guard Concepts and Administration manual, part no. A96653-02. regards, Carel-Jan At 15:54 8-12-03 -0800, you wrote: -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Jonathan Lewis INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
Re: Oracle Data Guard
Hi, Jonathan, I think your question is why I mentioned TDU, not just SDU, in my response to Guang's message. I admit I didn't give much thought and threw that in. Note:44694.1 says it's set to 32k by default and its adjustable range is 0 to 32k. Then the question is why Oracle chose the magical 32k. Would changing it to anything else yield any SQL*Net performance gain? It won't be too much extra work for Guang Mei to find out while he's experimenting with SDU. It's a little confusing when you say TDU is MTU, because, I think, the term MTU (Maximum transmission unit) is already used by network engineers to refer to the maximum number of bytes a data link layer frame can contain (1500 bytes for Ethernet e.g.). But I understand what you mean. Regarding a less than maximum SDU size, maybe it's useful if most of the SQL result is much less than 32k? Somebody can experiment and find out. Yong Huang Jonathan Lewis wrote: Can you clarify a couple of points for me. The SDU (session data unit) is presumably the packet size that the Oracle client and server want to pass back and forth - which is presumably the maximum size the one synchronous dialogue unit will be. The TDU (transport data unit) is presumably the predicted size of the transport maximum unit of data transfer (MTU). a) Why does Oracle need to know anything about the underlying transport mechanism ? b) If I set the SDU to the largest legal value (possibly 32K, perhaps 64K) the server task switch will occur after building and sending that packet - is there any good reason why I shouldn't do that. After all, if the transport simply accepts the 64K packet and gets it to the other end of the wire (not yet to the client session, just to the receiving transport layer) as rapidly as possible does it matter to Oracle whether the transport is using 1.5K or 8K packets. The fact that the transport layer doesn't have to work its packet synchronously means that some overheads have disappeared as far as Oracle is concerned. Regards Jonathan Lewis Hi, Guang, Look up SDU and TDU in Oracle documentation Network configuration. You set them in tnsnames.ora and listener.ora, not sqlnet.ora. protocol.ora allows you to modify some procotol-specific parameters. In addition, in your client application, you can choose a sensible array fetch size, such as arraysize in sqlplus (in fact, sqlplus arraysize changes more than just network data chunk size). You can't magically increase the network transfer rate by lowering network latency. But you can indirectly increase the rate by other means, such as buffering slightly more data in one chunk. Yong Huang __ Do you Yahoo!? New Yahoo! Photos - easier uploading and sharing. http://photos.yahoo.com/ -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Yong Huang INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
RE: Oracle Data Guard
Hi, Guang, Look up SDU and TDU in Oracle documentation Network configuration. You set them in tnsnames.ora and listener.ora, not sqlnet.ora. protocol.ora allows you to modify some procotol-specific parameters. In addition, in your client application, you can choose a sensible array fetch size, such as arraysize in sqlplus (in fact, sqlplus arraysize changes more than just network data chunk size). You can't magically increase the network transfer rate by lowering network latency. But you can indirectly increase the rate by other means, such as buffering slightly more data in one chunk. Yong Huang Guang Mei wrote: I have never worked on Network stuff. But is there any easy parameters we could set in sqlnet.ora so that we could increase the DB performance by increase the network transfer rate (without doing anything else)? BTW my sqlnet.ora (on a Sun Box) has only two lines: ... NAMES.DEFAULT_DOMAIN = incyte.com NAMES.DIRECTORY_PATH= (TNSNAMES, ONAMES, HOSTNAME) Guang __ Do you Yahoo!? New Yahoo! Photos - easier uploading and sharing. http://photos.yahoo.com/ -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Yong Huang INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
Re: Oracle Data Guard
Jumbo frames are new to me. The Ethernet Definitive Guide book says it was proposed by one vendor and adopted by several, so may not have good interoperability. But I wonder how much performance improvement there is by going from MTU 1500 with SDU 8k to MTU 8k with SDU 8k. I assume the lower the OSI level, the faster assembling and disassembling those protocol data units is done. Yong Huang With TCP over standard ethernet the maximum transfer unit (MTU) is about 1500 bytes, this means if you want to send 2000 bytes over network, you have to fragment it in 2 packets and send them separately. This means double packet headers, double latency etc. Jumbo frames is a capability of some Gbit ethernet cards which allow them to transfer about 9000 bytes in a single packet. SDU is session level transfer unit (session data unit). When you enable jumbo frames and set MTU/SDU to 8192 for example, you'll fit much more in single packet, thus increasing performance for larger transactions. __ Do you Yahoo!? New Yahoo! Photos - easier uploading and sharing. http://photos.yahoo.com/ -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Yong Huang INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
RE: Oracle Data Guard
Hi All, I am working on a similar project here. I am wondering if anyone in the list ever compared Oracle Data Guard with iReflect from Data Mirror. Please share your experience with us. Thanks, Kitty -Original Message- Sent: Monday, December 08, 2003 6:54 PM To: Multiple recipients of list ORACLE-L Hi Carel, What if 50% of tables doesn't have Primary/Unique keys, how it is going be with LSB then? Can you please explain more. with thanks, Vi --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 14:54 8-12-03 -0800, you wrote: Hi Carel, That is good help, can you please send me the pdf that you implemented there then. Was on its way already Tell me one thing I agree that we some times (rather most of the time ) generate less redo so we should be smooth. Can you tell me is there any releation between LSB and Primary keys, I read like LCR(logical Change Request) is based on Primary keys as It does not depends on Transaction at that time. Because LSB 'reverse engineers' SQL from the redolog info, it needs to get hold of the right rows. The rows get inserted/updated/deleted, and _a_ unique identification, not being the rowid, is required. So, every row needs to be uniquely identified. Have you implemnented LSB successfully? Yes, using a PSB / LSB combination for standby and reporting purposes respectively. with many thanks, Vi. --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 13:34 8-12-03 -0800, you wrote: Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. It all depends on the amount of redolog you generate. When that's pretty much, you waste some resources by transporting online/archived redologs you actually don't need. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). So, they don't like nologging operations like truncate, not even on the standby database? The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. LSB might work, but do not consider the option of failing over to it. Be aware that, altough in maximum protection mode your redolog arrives at the SB system within the transaction, it doesn't get applied there instantly. SQL Application takes place _after_ the log-switch on the Primary. When you take 10 minutes of redolog, and perform a logswitch, the SQL Apply process might even take longer than 10 minutes to complete processing of the redologfile. There is a risk that not every transaction arrives within 20 minutes at the LSB. So, your log-switching frequency and the amount of redo you generate per unit of time both play a major role in the refresh rate of the LSB. I'll send you the PDF of a DG Special I did in Kista a few months ago. Regards, Carel-Jan -- There will allwasy be another 10 last bugs -- Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc
RE: Oracle Data Guard
I have never worked on Network stuff. But is there any easy parameters we could set in sqlnet.ora so that we could increase the DB performance by increase the network transfer rate (without doing anything else)? BTW my sqlnet.ora (on a Sun Box) has only two lines: -- bash-2.03$ more sqlnet.ora # SQLNET.ORA Network Configuration File: /oracle/product/8.1.7/network/admin/sqlnet.ora # Generated by Oracle configuration tools. NAMES.DEFAULT_DOMAIN = incyte.com NAMES.DIRECTORY_PATH= (TNSNAMES, ONAMES, HOSTNAME) Guang -Original Message- Yong Huang Sent: Tuesday, December 09, 2003 2:59 PM To: Multiple recipients of list ORACLE-L Jumbo frames are new to me. The Ethernet Definitive Guide book says it was proposed by one vendor and adopted by several, so may not have good interoperability. But I wonder how much performance improvement there is by going from MTU 1500 with SDU 8k to MTU 8k with SDU 8k. I assume the lower the OSI level, the faster assembling and disassembling those protocol data units is done. Yong Huang With TCP over standard ethernet the maximum transfer unit (MTU) is about 1500 bytes, this means if you want to send 2000 bytes over network, you have to fragment it in 2 packets and send them separately. This means double packet headers, double latency etc. Jumbo frames is a capability of some Gbit ethernet cards which allow them to transfer about 9000 bytes in a single packet. SDU is session level transfer unit (session data unit). When you enable jumbo frames and set MTU/SDU to 8192 for example, you'll fit much more in single packet, thus increasing performance for larger transactions. __ -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Guang Mei INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
Re: Oracle Data Guard
Hi Vi, Rows NEED unique identification. So, if there are bunches of raw data with no unique identifier whatsoever (remember, rowid is not allowed) LSB can't generate a where-clause what row to update or delete on the SB database. It's generating SQL based on redolog info, and has to come up with an UPDATE table SET WHERE unique id = unique id. The unique id may be a multi-column key. There is an escape. Enabling supplemental logging can add extra info to do the unique identification, when no usable keys are available. This will cause some extra logging to be generated, there ain't no such thing as a free lunch. For detailed information read chapter 4.1.5 4.1.6 in the Oracle Data Guard Concepts and Administration manual, part no. A96653-02. regards, Carel-Jan At 15:54 8-12-03 -0800, you wrote: Hi Carel, What if 50% of tables doesn't have Primary/Unique keys, how it is going be with LSB then? Can you please explain more. with thanks, Vi - --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 14:54 8-12-03 -0800, you wrote: Hi Carel, That is good help, can you please send me the pdf that you implemented there then. Was on its way already Tell me one thing I agree that we some times (rather most of the time ) generate less redo so we should be smooth. Can you tell me is there any releation between LSB and Primary keys, I read like LCR(logical Change Request) is based on Primary keys as It does not depends on Transaction at that time. Because LSB 'reverse engineers' SQL from the redolog info, it needs to get hold of the right rows. The rows get inserted/updated/deleted, and _a_ unique identification, not being the rowid, is required. So, every row needs to be uniquely identified. Have you implemnented LSB successfully? Yes, using a PSB / LSB combination for standby and reporting purposes respectively. with many thanks, Vi. --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 13:34 8-12-03 -0800, you wrote: Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. It all depends on the amount of redolog you generate. When that's pretty much, you waste some resources by transporting online/archived redologs you actually don't need. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). So, they don't like nologging operations like truncate, not even on the standby database? The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. LSB might work, but do not consider the option of failing over to it. Be aware that, altough in maximum protection mode your redolog arrives at the SB system within the transaction, it doesn't get applied there instantly. SQL Application takes place _after_ the log-switch on the Primary. When you take 10 minutes of redolog, and perform a logswitch, the SQL Apply process might even take longer than 10 minutes to complete processing of the redologfile. There is a risk that not every transaction arrives within 20 minutes at the LSB. So, your log-switching frequency and the amount of redo you generate per unit of time both play a major role in the refresh rate of the LSB. I'll send you the PDF of a DG Special I did in Kista a few months ago. Regards, Carel-Jan -- There will allwasy be another 10 last bugs -- Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size
RE: Oracle Data Guard
Hi Kitty, Never heard of that, but I'm interested in your experiences, and the architecture of iReflect. The only advantage of Data Guard I can tell you is that it comes for free with your Oracle licences. Regards, Carel-Jan -- There will allways be another 10 last bugs -- At 12:14 9-12-03 -0800, you wrote: Hi All, I am working on a similar project here. I am wondering if anyone in the list ever compared Oracle Data Guard with iReflect from Data Mirror. Please share your experience with us. Thanks, Kitty -Original Message- Sent: Monday, December 08, 2003 6:54 PM To: Multiple recipients of list ORACLE-L Hi Carel, What if 50% of tables doesn't have Primary/Unique keys, how it is going be with LSB then? Can you please explain more. with thanks, Vi --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 14:54 8-12-03 -0800, you wrote: Hi Carel, That is good help, can you please send me the pdf that you implemented there then. Was on its way already Tell me one thing I agree that we some times (rather most of the time ) generate less redo so we should be smooth. Can you tell me is there any releation between LSB and Primary keys, I read like LCR(logical Change Request) is based on Primary keys as It does not depends on Transaction at that time. Because LSB 'reverse engineers' SQL from the redolog info, it needs to get hold of the right rows. The rows get inserted/updated/deleted, and _a_ unique identification, not being the rowid, is required. So, every row needs to be uniquely identified. Have you implemnented LSB successfully? Yes, using a PSB / LSB combination for standby and reporting purposes respectively. with many thanks, Vi. --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 13:34 8-12-03 -0800, you wrote: Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. It all depends on the amount of redolog you generate. When that's pretty much, you waste some resources by transporting online/archived redologs you actually don't need. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). So, they don't like nologging operations like truncate, not even on the standby database? The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. LSB might work, but do not consider the option of failing over to it. Be aware that, altough in maximum protection mode your redolog arrives at the SB system within the transaction, it doesn't get applied there instantly. SQL Application takes place _after_ the log-switch on the Primary. When you take 10 minutes of redolog, and perform a logswitch, the SQL Apply process might even take longer than 10 minutes to complete processing of the redologfile. There is a risk that not every transaction arrives within 20 minutes at the LSB. So, your log-switching frequency and the amount of redo you generate per unit of time both play a major role in the refresh rate of the LSB. I'll send you the PDF of a DG Special I did in Kista a few months ago. Regards, Carel-Jan -- There will allwasy be another 10 last bugs -- Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard
Oracle Data Guard
Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? 2) If yes then, is there any performance impact on Target/Source server database. 3) any drawbacks using Data Guard. with many thanks, Vijay. BT Yahoo! Broadband - Save £80 when you order online today. Hurry! Offer ends 21st December 2003. The way the internet was meant to be. http://uk.rd.yahoo.com/evt=21064/*http://btyahoo.yahoo.co.uk -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: =?iso-8859-1?q?Nalla=20Ravi?= INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
Re: Oracle Data Guard
Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't recommend you to use logical stby yet, it's more like a prototype currently, not exactly a working product. Tanel. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
Re: Oracle Data Guard
Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't recommend you to use logical stby yet, it's more like a prototype currently, not exactly a working product. Tanel. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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). BT Yahoo! Broadband - Save £80 when you order online today. Hurry! Offer ends 21st December 2003. The way the internet was meant to be. http://uk.rd.yahoo.com/evt=21064/*http://btyahoo.yahoo.co.uk -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: =?iso-8859-1?q?Nalla=20Ravi?= INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
RE: Oracle Data Guard
Hi, Tanel, enabling jumbo frames and SDU size if using Gbit ethernet, can you elaborate on this? -Original Message- Sent: Monday, December 08, 2003 1:34 PM To: Multiple recipients of list ORACLE-L Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't recommend you to use logical stby yet, it's more like a prototype currently, not exactly a working product. Tanel. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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). BT Yahoo! Broadband - Save £80 when you order online today. Hurry! Offer ends 21st December 2003. The way the internet was meant to be. http://uk.rd.yahoo.com/evt=21064/*http://btyahoo.yahoo.co.uk -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: =?iso-8859-1?q?Nalla=20Ravi?= INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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.net -- Author: Josh Collier INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
Re: Oracle Data Guard
Hi Vijay, I've done around 20 DG installations at different sites, using Linux, Solaris, AIX and W2K. High speed network isn't always what you need, low latency might be more important. I've set up a DG environment between Kuala Lumpur and Rotterdam, using a 128KB line. This wasn't for standby reasins, but was about database consolidation. As long as the daily Archive Log could be done, everything was OK. For maximum availability/protection mode you get synchronous writes. Latency counts very much. See also DocId 233650.1 on Metalink, Titled 9i Data Guard Complete Reference. It actually contains a list of interesting Data Guard publications on metalink. Force logging mode is required for Logical Standby. I agree with Tanel, LSB isn't for production yet. However, in some situations it might be usefull for a reporting database, as long as you do not rely on the standby part. Create a PSB for that purpose. On one site, with a long distance PSB ( 100 km between datacenters), with requirements about Zero Data Loss, I advised to setup a PSB/maximum availabilty mode locally, and a remote PSB in max. performance mode. This minimizes the chance of data loss, and affects performance as little as possible. Regards, Carel-Jan -- There will allwasy be another 10 last bugs -- At 07:19 8-12-03 -0800, you wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't recommend you to use logical stby yet, it's more like a prototype currently, not exactly a working product. Tanel. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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.net -- Author: Carel-Jan Engel INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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).
Re: Oracle Data Guard
Comments inline At 13:34 8-12-03 -0800, you wrote: Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. It all depends on the amount of redolog you generate. When that's pretty much, you waste some resources by transporting online/archived redologs you actually don't need. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). So, they don't like nologging operations like truncate, not even on the standby database? The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. LSB might work, but do not consider the option of failing over to it. Be aware that, altough in maximum protection mode your redolog arrives at the SB system within the transaction, it doesn't get applied there instantly. SQL Application takes place _after_ the log-switch on the Primary. When you take 10 minutes of redolog, and perform a logswitch, the SQL Apply process might even take longer than 10 minutes to complete processing of the redologfile. There is a risk that not every transaction arrives within 20 minutes at the LSB. So, your log-switching frequency and the amount of redo you generate per unit of time both play a major role in the refresh rate of the LSB. I'll send you the PDF of a DG Special I did in Kista a few months ago. Regards, Carel-Jan -- There will allwasy be another 10 last bugs -- Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't recommend you to use logical stby yet, it's more like a prototype currently, not exactly a working product. Tanel. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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). BT Yahoo! Broadband - Save £80 when you order online today. Hurry! Offer ends 21st December 2003. The way the internet was meant to be. http://uk.rd.yahoo.com/evt=21064/*http://btyahoo.yahoo.co.uk -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: =?iso-8859-1?q?Nalla=20Ravi?= INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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.net -- Author: Carel-Jan Engel
Re: Oracle Data Guard
With TCP over standard ethernet the maximum transfer unit (MTU) is about 1500 bytes, this means if you want to send 2000 bytes over network, you have to fragment it in 2 packets and send them separately. This means double packet headers, double latency etc. Jumbo frames is a capability of some Gbit ethernet cards which allow them to transfer about 9000 bytes in a single packet. SDU is session level transfer unit (session data unit). When you enable jumbo frames and set MTU/SDU to 8192 for example, you'll fit much more in single packet, thus increasing performance for larger transactions. Also make sure that TCP.NODELAY is unset in your sqlnet.ora or is set to true. More info can be found from: http://download-west.oracle.com/docs/cd/B10501_01/server.920/a96653/troubleshooting.htm#635905 Tanel. - Original Message - To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED] Sent: Monday, December 08, 2003 11:49 PM Hi, Tanel, enabling jumbo frames and SDU size if using Gbit ethernet, can you elaborate on this? -Original Message- Sent: Monday, December 08, 2003 1:34 PM To: Multiple recipients of list ORACLE-L Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't recommend you to use logical stby yet, it's more like a prototype currently, not exactly a working product. Tanel. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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). BT Yahoo! Broadband - Save £80 when you order online today. Hurry! Offer ends 21st December 2003. The way the internet was meant to be. http://uk.rd.yahoo.com/evt=21064/*http://btyahoo.yahoo.co.uk -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: =?iso-8859-1?q?Nalla=20Ravi?= INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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
Re: Oracle Data Guard
Hi Carel, That is good help, can you please send me the pdf that you implemented there then. Tell me one thing I agree that we some times (rather most of the time ) generate less redo so we should be smooth. Can you tell me is there any releation between LSB and Primary keys, I read like LCR(logical Change Request) is based on Primary keys as It does not depends on Transaction at that time. Have you implemnented LSB successfully? with many thanks, Vi. --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 13:34 8-12-03 -0800, you wrote: Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. It all depends on the amount of redolog you generate. When that's pretty much, you waste some resources by transporting online/archived redologs you actually don't need. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). So, they don't like nologging operations like truncate, not even on the standby database? The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. LSB might work, but do not consider the option of failing over to it. Be aware that, altough in maximum protection mode your redolog arrives at the SB system within the transaction, it doesn't get applied there instantly. SQL Application takes place _after_ the log-switch on the Primary. When you take 10 minutes of redolog, and perform a logswitch, the SQL Apply process might even take longer than 10 minutes to complete processing of the redologfile. There is a risk that not every transaction arrives within 20 minutes at the LSB. So, your log-switching frequency and the amount of redo you generate per unit of time both play a major role in the refresh rate of the LSB. I'll send you the PDF of a DG Special I did in Kista a few months ago. Regards, Carel-Jan -- There will allwasy be another 10 last bugs -- Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't recommend you to use logical stby yet, it's more like a prototype currently, not exactly a working product. Tanel. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (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). BT Yahoo! Broadband - Save £80 when you order online today. Hurry! Offer ends 21st December 2003. The way the internet was meant to be.
Re: Oracle Data Guard
Comments inline At 14:54 8-12-03 -0800, you wrote: Hi Carel, That is good help, can you please send me the pdf that you implemented there then. Was on its way already Tell me one thing I agree that we some times (rather most of the time ) generate less redo so we should be smooth. Can you tell me is there any releation between LSB and Primary keys, I read like LCR(logical Change Request) is based on Primary keys as It does not depends on Transaction at that time. Because LSB 'reverse engineers' SQL from the redolog info, it needs to get hold of the right rows. The rows get inserted/updated/deleted, and _a_ unique identification, not being the rowid, is required. So, every row needs to be uniquely identified. Have you implemnented LSB successfully? Yes, using a PSB / LSB combination for standby and reporting purposes respectively. with many thanks, Vi. --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 13:34 8-12-03 -0800, you wrote: Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. It all depends on the amount of redolog you generate. When that's pretty much, you waste some resources by transporting online/archived redologs you actually don't need. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). So, they don't like nologging operations like truncate, not even on the standby database? The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. LSB might work, but do not consider the option of failing over to it. Be aware that, altough in maximum protection mode your redolog arrives at the SB system within the transaction, it doesn't get applied there instantly. SQL Application takes place _after_ the log-switch on the Primary. When you take 10 minutes of redolog, and perform a logswitch, the SQL Apply process might even take longer than 10 minutes to complete processing of the redologfile. There is a risk that not every transaction arrives within 20 minutes at the LSB. So, your log-switching frequency and the amount of redo you generate per unit of time both play a major role in the refresh rate of the LSB. I'll send you the PDF of a DG Special I did in Kista a few months ago. Regards, Carel-Jan -- There will allwasy be another 10 last bugs -- Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't recommend you to use logical stby yet, it's more like a prototype currently, not exactly a working product. Tanel. -- Please see the official ORACLE-L FAQ: http://www.orafaq.net -- Author: Tanel Poder INET: [EMAIL PROTECTED] Fat City Network Services-- 858-538-5051 http://www.fatcity.com San Diego, California-- Mailing list and web hosting services - To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing
Re: Oracle Data Guard
Hi Carel, What if 50% of tables doesn't have Primary/Unique keys, how it is going be with LSB then? Can you please explain more. with thanks, Vi --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 14:54 8-12-03 -0800, you wrote: Hi Carel, That is good help, can you please send me the pdf that you implemented there then. Was on its way already Tell me one thing I agree that we some times (rather most of the time ) generate less redo so we should be smooth. Can you tell me is there any releation between LSB and Primary keys, I read like LCR(logical Change Request) is based on Primary keys as It does not depends on Transaction at that time. Because LSB 'reverse engineers' SQL from the redolog info, it needs to get hold of the right rows. The rows get inserted/updated/deleted, and _a_ unique identification, not being the rowid, is required. So, every row needs to be uniquely identified. Have you implemnented LSB successfully? Yes, using a PSB / LSB combination for standby and reporting purposes respectively. with many thanks, Vi. --- Carel-Jan Engel [EMAIL PROTECTED] wrote: Comments inline At 13:34 8-12-03 -0800, you wrote: Hi Tanel, Much appreciated, The fact is I am interested in Logical standby rather than physical. Our 30-50% of our Production data needs to be replicated to another database and where they will have their processing and batches. It all depends on the amount of redolog you generate. When that's pretty much, you waste some resources by transporting online/archived redologs you actually don't need. Now We didn't go to Snapshot because It is on multiple tables (where we didnot have PK's and many tables) and due to performance issue I didn't want to use Snapshots (they did not want any tables to be truncate before being loaded even via snapshots). So, they don't like nologging operations like truncate, not even on the standby database? The best option I think is Logical Standby Database. Or Can you please suggest me any other means. Replication should be quicker like once in every 20 minutes, Even Transportable tablespacs does not work here since they need all tables to 24*7. LSB might work, but do not consider the option of failing over to it. Be aware that, altough in maximum protection mode your redolog arrives at the SB system within the transaction, it doesn't get applied there instantly. SQL Application takes place _after_ the log-switch on the Primary. When you take 10 minutes of redolog, and perform a logswitch, the SQL Apply process might even take longer than 10 minutes to complete processing of the redologfile. There is a risk that not every transaction arrives within 20 minutes at the LSB. So, your log-switching frequency and the amount of redo you generate per unit of time both play a major role in the refresh rate of the LSB. I'll send you the PDF of a DG Special I did in Kista a few months ago. Regards, Carel-Jan -- There will allwasy be another 10 last bugs -- Any suggestion would be more helpful. with thanks, Vi. --- Tanel Poder [EMAIL PROTECTED] wrote: Hi All, can any one let me know kindly the following info. 1) Has any one used the Oracle 9i Data Guard? Yes, physical standby and successfully. 2) If yes then, is there any performance impact on Target/Source server database. Your database has to be in archivelog mode, but when you are thinking such solutions as DG, then you probably are already running archivelog anyway. If you run in maximum protection or maximum availability, yes there is. The impact depends mainly on network connection between primary and standby(s) and the speed of redolog disks. You could tune these by using faster network, enabling jumbo frames and SDU size if using Gbit ethernet, also setting lgwr and log apply processes priority higher than others. 3) any drawbacks using Data Guard. You should set your database or critical tablespaces to force logging mode in order to transfer all changes to standby in physical standby. That means, performance improvements which take advantage of nologging operations (such insert append nologging etc), will not run that fast anymore. In logical standby, I think there's no such requirement, but I don't === message truncated === BT Yahoo! Broadband - Save £80 when you order online today. Hurry! Offer ends 21st December
Oracle Data Guard
Hi. I'm looking into implementing a read-only stand-by database. I'd like to have it read-only for periodic reporting during the day. Is this possible under 8.1.7, or 9i? Are there any drawbacks to having the stand-by database opened for read-only? I would have two goals in mind: 1. A stand-by, read-only reporting db. 2. Easy switch-over to the stand-by. Are these two goals compatible? I can't find any documention on Data Guard for 8.1.7, as the links on Technet are broken. Thanks Jeff -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Jeff Wiegard INET: [EMAIL PROTECTED] 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: [EMAIL PROTECTED] (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).
RE: Oracle Data Guard
Jeff - With 8.1.7, I believe that you are limited to read-only or standby, but you can't have both simultaneously. It is either in recovery mode accepting redo logs from production or open and allowing read-only. I believe that with 9i the options are much more flexible. Read up on the Oracle9i Real Application Clusters. Dennis Williams DBA Lifetouch, Inc. [EMAIL PROTECTED] -Original Message- Sent: Thursday, November 01, 2001 1:20 PM To: Multiple recipients of list ORACLE-L Hi. I'm looking into implementing a read-only stand-by database. I'd like to have it read-only for periodic reporting during the day. Is this possible under 8.1.7, or 9i? Are there any drawbacks to having the stand-by database opened for read-only? I would have two goals in mind: 1. A stand-by, read-only reporting db. 2. Easy switch-over to the stand-by. Are these two goals compatible? I can't find any documention on Data Guard for 8.1.7, as the links on Technet are broken. Thanks Jeff -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Jeff Wiegard INET: [EMAIL PROTECTED] 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: [EMAIL PROTECTED] (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: DENNIS WILLIAMS INET: [EMAIL PROTECTED] 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: [EMAIL PROTECTED] (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).
Oracle Data Guard
Hi Gang, Can anyonepoint me to some documentation for this product? Technet seems pretty limited. TIA, -Rocky __ Do You Yahoo!? Get email at your own domain with Yahoo! Mail. http://personal.mail.yahoo.com/ -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Rocky Welch INET: [EMAIL PROTECTED] 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: [EMAIL PROTECTED] (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).
