Distributed database

Similar Questions

• How to activate the validation to phase 2 in oracle11g (for oracle distributed databases)

  Hi all
  
  We have three oracle servers(11g) and we need to update the data to a server on the remaining servers. You want to know, is "two phase commit" automatically activated/installed in 11g or must be allowed. If we need to enable it then what are the steps all required them for the same.
  
  Thanks and greetings
  Ravi

  It is always nice to see people think they are so unique that their question has already been answered in the documentation.

  Yes, the two-phase commit is enabled.
  No, depending on what you want, automatically update the records of one server to other servers is probably a very bad idea, because Oracle has documented of the facilities for this.
  You have among other things
  -Materialized views
  -Oracle Streams

  Certainly much better than your approach scale these solutions, because they are asynchronous. In your approach to the master server will slow down due to the update of the slaves.
  Approaches need to be avoided in distributed databases 'push' and 'pull' to focus on approaches. That's exactly what materialized views and streams make.

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

• Trends of database management systems

  Hello
  
  Y at - it an official document that attempts to predict the trends of database management system? As cloud computing? Distributed database systems?
  
  Thank you.

  What does "official document" for you?

  For example, you can go look at oracle Open World keynotes or sessions for various takes on the future of databases. You can buy tons of market from companies like Gartner research. Many technology companies, technology industry journalists, etc. spend countless hours writing articles that speculate on where various technologies will in the future.

  Oracle Corporation does not generally 'official' predictions about the future of the public (this operation would be extremely complicated securities regulators). Discussions of what might or might not be in the next version of the database must be carefully controlled and be accompanied by discussions of 'safe harbor' repeating that nothing in presentations is a promise to implement a specific feature and that nobody should be based on new information provided for any purpose.

  Justin

• referential integrity distributed

  Hi, I have a question. If I'm on the wrong forum please direct me because I couldn't find a specific forum for distrituted databases and it can be a matter of application development.
  
  I implement a simple distributed database application. At this point I fear that with the creation of two tables of each on a different server. For the simplicity I will be their name table1 and table2. This is the scheme if the tables were not distributed and were located on the same server.
  
  CREATE TABLE table1)
  PRIMARY KEY of the CONSTRAINT of the NUMBER of table1_pk table1_id
  );
  
  CREATE TABLE table2)
  NUMBER of table2_id
  NUMBER of table1_id
  CONSTRAINTtable2_pk PRIMARY KEY (table2_id, table1_id),
  CONSTRAINT table2_fk FOREIGN KEY (table1_id) REFERENCES table1 (table1_id)
  );
  
  I talked about literature in the database Oracle® 10g Release 2 Administrator's Guide (10.2) and the Oracle database applications developer's Guide. I know I can't use declarative referential integrity, but how can I distribute the tables so that I can be sure that when I insert a row in table2, I can check to see if there is a corresponding line in table 1 for table1_id?

  Actually, I expect that you will review your architecture.

  It makes no sense to search one matching row in table1 a remote database when the data is inserted in table2 in a local database. The best scenario would be that each insertion in table2 would be overloading a network round trip plus costs of querying the table in the remote database. If the network is down or if one of the two databases is down, the application fails, that beats in general the object of a distributed application. In addition, there will be all sorts of concurrency issues (i.e. I remove a line, but before I commit to you ask the line, to see there and insert a child line.) I agree, leaving your orphan line).

  Assuming that you really do need a distributed architecture, you want to replicate table1 to the nodes the and remote. Then, you would report to referential integrity between your local copy of table1 and table2 (as well as between table1 and table2 on the remote database, assuming you wanted to table2 data available here as well). Your replication process (preferably using stream but potentially use the multi master materialized views rather) will then be coded to process the errors due to the asynchronous nature of replication (that is to notice that database1 deleted a parent row that you just inserted a child line for and solve the conflict in an appropriate manner).

  Justin

• Oracle Forms 10 g for distributed environment

  Hi all
  
  I tried searching through the documentation for Oracle Forms, but I can't find anything about it either related to Oracle Forms development for a distributed database. Someone at - it links to documents or web sites that could help me? I would really appreciate it.
  
  Kind regards
  John

  In this case, your question is probably best answered on the forum of the database. What forms is it sees tables according to the privileges of the logged in user has. It should be transparent to the forms if, for example, database links are being physically stored reference elsewhere tables user...

  Grant

• Connection app mail to email to colleagues

  Hello

  in my work, we use IBM Lotus Notes as an e-mail client. There is an interesting feature. Each user can set other users to see their mailbox. So when this person is going for vacation, the other person can easily connect to his mailbox and can see all the messages. So when our customers send an E-Mail to someone who does not work at this time here, he always gets a response. There is the possibility to implement "office" message but it's not good in the newspaper to another account.

  Is somehow possible something similar in Mail.app?

  Thanks for any idea

  Mail application is just an email client. You can configure some email (POP, IMAP, or Exchange) accounts you have the credentials for.

  Lotus Notes is off in his own little world... Notes is not an e-mail program, Notes is a distributed database. It may look like a platform of messaging for you, but it's not...

  Talk to your it Department of your options.

• Need a domain name?

  Hello

  in recent weeks, trying to set up a database of the 12 Oracle and Oracle Enterprise Manager on a Microsoft Windows Server 2008 R2 virtual computer. The last time I had some problems connecting to the PDB, so I deleted the CBD and started around. Then, I couldn't even connect to the CBD.

  I chose an easy global name: 'WS2014' without any domain name. Everything I tried, I could not connect. So I suspect that the domain name is necessary, isn't it?

  Now, I created a false domain name in my hosts file:

  192.168.73.2	local.com
  192.168.73.128	ORCL.local.com	ORCL

  The new global name now is orcl.local.com, the name of the service of the CDB is also orcl.local.com (the installer set on its own to this value) and the PDB is pdborcl.local.com.

  After the opening of the PDB (alter connectable to all database open; why I have to explicitly open them?) I can connect to the CBD and the PDB as well.

  So my questions as stated above:

  1. is the domain name needed, although it is only a fake that I put in the hosts file?
  2. What could be the problem on my last setup when I chose WS2014 as the name of name service / overall?
  3. Why should I run alter database connectable to all open; explicitly?

  Thanks in advance!

  in recent weeks, trying to set up a database of the 12 Oracle and Oracle Enterprise Manager on a Microsoft Windows Server 2008 R2 virtual computer. The last time I had some problems connecting to the PDB

  Well - we cannot help you at the level of the "problems you might have had since you do not" post.

  The most common problem with 12 c is that people didn't RTFM before you begin to install and you try to use the database.

  If you have not already done, so you must read and understand the chapters 17 and 18 of the doc of Concepts on the new multitenancy of the 12 c.

  http://docs.Oracle.com/CD/E16655_01/server.121/e17633/part_consol.htm#CHDGDBHJ

  These chapters introduce the new architecture and BEGIN (and I emphasize the word 'start') the process of clarification on the differences again. They talk about some of the differences between users the and common.

  The next most common 'number' is trying to connect to a PDB file that has not been opened. The START command does NOT open the PDB files. If you want to pdb open for you is obliged to open them manually after the boot or create a trigger AFTER STARTUP that runs the alter and opens to them.

  Third place 'question' trying to connect ot a PDB file that does not have the appropriate entry in the TNSNAMES file. ORA or try to use the SID instead of SERVICE_NAME. The 12 c DB REQUIRES service_name, not MSN.

  The fourth most common 'question' is trying to create a local user in the CBD. Local users cannot be created in the CBD - only in the PDB files. But then the people who have not read these chapters two doc don't know the difference between a common user and a local user so they get exceptions, they do not understand. Save yourself a LOT of trouble and read these two chapters.

  I chose an easy global name: 'WS2014' without any domain name. Everything I tried, I could not connect. So I suspect that the domain name is necessary, isn't it?

  No - it is NOT necessary in 12 c and hasn't ' st was necessary in any previous version. It is also in the documentation.

  See "determine the global database name"

  https://docs.Oracle.com/database/121/Admin/create.htm#i1013998

  The global database name includes the name specified by the local database user and location of the database in a network structure. The DB_NAME initialization parameter determines the local name of the name of database component and the DB_DOMAIN optional parameter indicates the domain (logical location) within a network structure. The combination of settings for these two parameters must form a unique database name within a network.

  DB_DOMAIN initialization parameter

  DB_DOMAINis a text string that specifies the network domain where the database is created. If the database you are about to create will never be part of a distributed database system, and then pay special attention to this initialization parameter before the creation of the database. This parameter is optional.

  If you use the DB_DOMAIN parameter, then the overall DB name includes the value that you specify.

  After the opening of the PDB (alter connectable to all database open; why I have to explicitly open them?) I can connect to the CBD and the PDB as well.

  PDB does NOT open by default. You need to manually open them, or you create a trigger AFTER the opening of SESSION that issues an ALTER statement to open the PDB files that you want to open.

  If Oracle has just opened all the PDBs default (or rather "tried" to open) you don't have control over those who had opened. Oracle would even try to open those that you have intentionally shutdown. What sense would it make?

  A PDB file is a database. Like all databases - you want it opened, and then open.

  1. is the domain name needed, although it is only a fake that I put in the hosts file?
  2. What could be the problem on my last setup when I chose WS2014 as the name of name service / overall?
  3. Why should I run alter database connectable to all open; explicitly?

  #1-No - see above reference doc

  #2-who knows - we cannot help you if you do not post the problem and the necessary info

  #3-already answered - if you want to open a database, ALL database, you must open. It's no different for 12 c for earlier versions of DB.

  I STRONGLY suggest that you take a break and RTFM - in particular chapters 17 and 18 of the concepts doc.

  If you do not you will display the questions again about 'where are default patterns', "why can I not connect to the PDB" or "why can I not create a user named User1.

• DB_DOMAIN

  Hello
  
  What are the effects of the definition of the DB_DOMAIN = all parameter?
  
  Thanks in advance
  KSG

  I just want to know if DB_DOMAIN = world is defined as means?

  It's a parameter by default if there is no field of network defined by you. As you would read in the links of the RFSO, the db_domain will be valid in a distributed database environment. So if you encounter a network defined for your company domain, and then set the db_domain to the default.

  Aman...

• How to value of primary key on insertions of autogen

  I'm under 4.0. I am writing an app that manages the inventory of distributed database information. All my tables have primary keys digital substitute. I created several combinations/report form. According the report, I can choose to create a new instance of an object; Server for example. When I click on the ' Add Server > ' button, the details of the page opens for me to create the new entry. On this page, I need to have the autogenerated Server_ID. I think something as simple as "Select (max (server_id) + 1) server ' should do. In fact, it seems that it will work because the field is correctly completed. But, when I click on the button 'Create' on this page I get an error that says the Server_ID sucks. Looks like that the generated value is not what makes the INSERT statement.
  
  On the setting Page, I together Server_ID in view only mode and in the Source section selected attribute that I "only when there is zero current in the session state value" and "SQL Query (single return value)" selected with ' Select (max (server_id) + 1) Server "in the field"Source value or expression.

  DO NOT USE the ' Select (max (server_id) + 1) server to assign a new value of PK...» He has so many questions, I don't want to start with this... If you need to get the PK BEFORE all of the data on the form, then have a calaculation on e electronic form that is done before header where you select the NEXT value in the sequence.

  Thank you

  Tony Miller
  Webster, TX

• Is something wrong with the forum?

  I see threads and replies endangered.

  It seems that it has something to do with the fact that the forum is actually served by a number of different servers with a distributed database.

  I guess I'm lucky, it seems pretty solid for me today.  I probably was assigned to a server without a database problem.

  -Christmas

• API in EBS

  Hi hussein.
  
  We are implementing EBS R12 on our new client. They have other cultivated host applications that need
  interface with the EBS, which means that they will exchange data. Homegrown apps are developed from SQL Server,
  FoxPro and access.
  
  How to make the exchange of data between these applications 'in House' and the EBS? My boss told me to use 'API '.
  What are the "api"? Are these Oracle PL/SQL procedures and functions? Have you tried to use?
  Can you give me Metalink notes/docs re: api?
  Have you heard of homogeneous/heterogeneous connectivity?
  
  
  Thank you very much
  
  Mrs. k

  Hello

  How to make the exchange of data between these applications 'in House' and the EBS? My boss told me to use 'API '.
  What are the "api"? Are these Oracle PL/SQL procedures and functions? Have you tried to use?
  Can you give me Metalink notes/docs re: api?

  Please see the links following/docs:

  The API Oracle world
  http://www.oracleappshub.com/EBS-suite/the-world-of-Oracle-API/

  Oracle applications API
  Re: Oracle applications API

  Note: 166324.1 - which is an Application Programming Interface (API)?
  https://metalink2.Oracle.com/MetaLink/PLSQL/ml2_documents.showDocument?p_database_id=not&P_ID=166324.1

  Oracle integration repository
  http://IREP.Oracle.com

  Have you heard of homogeneous/heterogeneous connectivity?

  Check out this link.

  Distributed database architecture
  http://download.Oracle.com/docs/CD/B28359_01/server.111/b28310/ds_concepts001.htm

  Kind regards
  Hussein

• level of Rollback statement and 2PC/2PL

  Mr President
  
  I'm very big thing back to oracle db but confused here:
  
  Level in oracle Rollback statement says:
  
  (1) If a single LMD fails while executing this single statement is rolled back.
  
  (2) the oracle server implements a default backup point.
  
  (3) all other changes are preserved.
  
  (4) the user has to interrupt transactions explicitly by executing a statement commit or rollback.
  
  issues related to the:
  
  ............. (1) assumes that I have a transaction T1, T2... TN.
  
  Obviously T1 starts with begin and ends to commit / rollback. It may look like:

   
  begins
  .............
  some DML statements
  
  commit / rollback.

  the same for the T2, T3,..., TN.
  
  Let say for transaction Ti a DML said update statement fails for bad network link.what will happen next?
  
  If my interpretation is not bad:
  
  A lock will run on the resources.and she will invest in the last commit point.
  
  but, according to the statement (what our faculty of WDP o interpreted) level restoration it will commit to the last well played DML when the central database server crashed eg.what occur if you restart your computer during an operation to copy - paste .If it's true how oracle ensures consistency? because the other site that can not get network connection, won't the other transaction state.
  
  Please comment on this and clear my doubt.
  
  Edited by: mango_boy, 1 Sep 2009 01:15
  edited the title as it was with a silly mistake ;-). i admit that I should be more careful for this terminology
  
  Published by: mango_boy on Sep 4, 2009 12:50
  (my review of teachers is being updated)

  A DDL statement cannot be part of a transaction (or, more precisely, before 11 g R2, issuing a DDL statement will be automatically and impliictly validate the previous transaction, starts a new transaction and validate this new transaction if it is successful). Since the DDL changes the structure of the database, it is exceptionally difficult to do in a transactional manner. You cannot restore a DDL operation (although Oracle can if the DDL statement fails). This may change to some extent in 11g R2, some of the new arrivals appear to allow several editions of the code runs in the same time, potentially changing DDL transaction semantics. It'll be a few months before the documentation is available on this feature, however.

  As you say errors in (this scenario) is not related to the error DDBMS, and these errors shouldn't be a syntactic error, if I am mistaken please clear my confusion) then what are the types of error in the database can cope?

  If update3 is not intended to be distributed operations (I guess that DDBMS is expected to mean "distributed database management system"), I guess it could be a SQL syntax error, but which would cause the statement fails before that she had even started, which rather defeats the purpose of roll back to a save point, because there would be no changes to roll back. In addition, a syntax error shouldn't be possible in a production application - if a SQL statement in a procedure has an error, the entire procedure may not compile. More commonly and more interesting for what you ask, would be the case where the update statement updates of 100 lines and then generated a line 101 update constraint violation.

  Justin

• precompute_subquery Council

  What exactly does the indicator precompute_subquery don't?
  Is the different behavior when the work through distributed databases of subquery data comes from the remote database?

  I have written a post internal and detailed about this behavior:

  [http://blog.tanelpoder.com/2009/01/23/multipart-cursor-subexecution-and-precompute_subquery-hint/]

• Set global_names = false in oracle

  Hello
  I have a doubt about the place where to set the global_names parameter. I mean on the remote database or the database when you go to...
  
  For ex: there are 2 databases on 2 different servers... say db1 and db2
  
  I created a llink from db1 to db2 database... .and the link is created...
  
  but when I try to access the data through the link does not work... it gives an error that the name of the link must be the same as the database. This happens because global_names = true...
  
  So I created the link with the name db2 database... and now I am able to access the remote database data.
  But because of that... .i can't keep changing the name of the db link in the code everywhere... it is not possible
  
  My question is: where to put this setting?
  Hope that's clear
  Thank you

  global_names must be set to the original database server, the server that holds the link.
  I recommend that you change your code and AGAINST the creation of global_names to false.
  He will break all replication Oracle facilities and related functions probably more distributed database.

  -----
  Sybrand Bakker
  Senior Oracle DBA

• narrowing of the shared pool

  Hi friends, when you browse the documentation I read that "internal adjustment algorithm isn't shrink shared pool when you use EAMA. I want to know if that's all the time or for specific cases. If for all the time, then it can lead to performance and more problem sql, pl/sql etc info will reside.
  
  Please explain to me. Thanks in advance

  Hello
  Yes in the case of the distributed databases, using pfile we can statup dbs with client side pfile as well. See here,
  The location of the file in distributed environments initialization parameter
  The client that allows you to access the database must be able to read the settings of the base initialization file to start instance database. Therefore, always store a file parameter on the computer running the client.

  In undistributed treatment facilities, the computer is running Oracle and the customer. This computer already has the settings file stored on one of its hard drives. In distributed processing facilities, however, local client workstations may administer a database stored on a remote computer. In this configuration, the local client computers must each store a copy of the settings for the corresponding databases file.
  http://download-West.Oracle.com/docs/CD/A87860_01/doc/server.817/a76956/start.htm#6439
  HTH
  Aman...