|
|
[01.00] Where is the latest version of this FAQ?
[01.01] How about a summary of what DataBee does?
[01.02] How can I contact the publishers of DataBee?
[01.03] I have an idea for a new/better feature - do you want to hear about it?
[01.04] Where can I find out more about how DataBee works?
[01.05] Can DataBee create a subset of an Oracle Applications instance?
Version and Operating System Questions
[02.01] What is the current release version of DataBee?
[02.02] What are the PC system requirements for DataBee?
[02.03] What are the server side requirements for DataBee?
[02.04] What server side operating systems does DataBee support?
[02.05] What Oracle database versions does DataBee support?
[02.06] What version of DataBee am I currently using?
[02.07] How can I see or edit my License key?
[02.08] How do I upgrade DataBee?
'How do I' and 'Can I' Type Questions
[03.01] How big can a DataBee Extracted Set File get?
[03.02] Can I share my Extracted Set File with other users?
[03.03] Can I put my extraction directory on a Network Drive?
[03.04] Can a table have more than one WHERE Clause rule?
[03.05] Can a table have more than one rule pointing out of it?
[03.06] Can a table have more than one rule pointing into it?
[03.07] Can a table have a rule pointing at itself?
[03.08] Can a table have more than one rule pointing at another table based on different join conditions?
[03.09] Can a table have more than one rule pointing at another table based on the same join condition?
[03.10] If table A is the Parent of table B in one rule can it be the Child of table B in another rule?
[03.11] Why does the Set Designer sometimes not know what columns a table has?
[03.12] The rest of the rules extract really fast. Why is one so slow?
[03.13] How do I change the directory in which the log files are placed?
[03.14] Do I have to have a plan in order to create an extraction set?
[03.15] My database has no foreign keys, can DataBee still extract a subset?
[03.16] Can I create a subset on a earlier version of Oracle than the source database?
[03.17] Can I create a referentially correct set of data from tables distributed between more than one schema?
[03.18] What do the letters FP mean in the extracted row count columns?
[03.19] Can DataBee extract and load tables containing user defined column types or LONGS?
Questions about What DataBee Is and Does
[01.00] Where is the latest version of this FAQ?
The latest version of the DataBee FAQ can be found on the DataBee web site at http://www.DataBee.com/dbfaq.htm[01.01] How about a summary of what DataBee does?The basic function of DataBee is to construct smaller Oracle databases which are composed of a subset of the rows contained in a larger Oracle database. This is a common requirement as it is quite undesirable (in terms of cost, time and performance) to give developers, testers and trainers full size copies of databases to work on. DataBee consists of four applications each of which is dedicated to a single function.[01.02] How can I contact the publishers of DataBee?
- The Set Designer application finds, edits and collects extraction rules together into a group known as an extraction set.
- The Set Extractor application uses the extraction set to identify the rows in the source database that will be required to build the subset. The identifiers for these rows are placed into a binary file known as an Extracted Set File.
- The Set Loader application uses the Extracted Set File to load the destination database. It will also handle the enabling of the relevant constraints and triggers.
- The Set Planner application is used to make decisions about the tables which will or will not be included in an extraction set. As a tool it defines the components of the extraction set and serves to bring into focus the exact requirements of the subset database.
- The freeware DDL Extract Wizard software, also published by Net 2000 Ltd., is used to create the empty structure (tables, packages, synonyms & etc) of the subset database.
- Another Net 2000 Ltd. software product called The Data Masker can be used to sanitize and scrub data. The Data Masker is designed to protect sensitive information in test databases by replacing it with realistic looking but false information.
DataBee is published by Net 2000 Ltd. Any email to DataBee@DataBee.com will get to us. Our full contact details are on this page.[01.03] I have an idea for a new/better feature - do you want to hear about it?For sure we do. Please E-mail DataBee@DataBee.com.[01.04] Where can I find out more about how DataBee works?Ah... you mean all of the technical internal workings. Chapter 2 of the DataBee Quick Start Guide contains a detailed explanation and the Test Database Generation and Management whitepaper contains a theoretical overview.[01.05] Can DataBee create a subset of an Oracle Applications instance?Yes. In fact, we offer a pre-prepared "black box" solution which will need only minimal modifications for your site specific customizations. This means you can create small, referentially correct, test and development Oracle Applications instances quickly and easily. The DataBee Oracle Applications functionality is not available for download from the website. Please have a look at our DataBee Oracle Applications page or email us at DataBee@DataBee.com for more information.
[02.01] What is the current release version of DataBee? Version and Operating System Questions
The current production release of the DataBee is v03.06[02.02] What are the PC system requirements for DataBee?DataBee works on Windows 98, NT4, Windows 2000, and Windows XP. All that is required is that the Oracle supplied SQL*Net v8 (or greater) networking software is installed on the PC. There are no really large hardware requirements - any reasonably new Pentium PC with more than 64Mb of memory should be sufficient. A reasonably large monitor is desirable (but not essential) and the screen resolution should be 800x600 or greater.[02.03] What are the server side requirements for DataBee?The DataBee software itself takes up very little space - 3 Mb of disk space should be more than sufficient. The Extracted Set Files can get fairly large. You might need to reserve 40 to 70Mb for them - depending on the number you wish to keep on hand and the number of rows extracted in each one.
Not very much. DataBee requires one small temporary table to be created in the source Oracle schemas. This table can be deleted after the extract completes and will never contain more than 1000 rows. No other components are needed. DataBee requires no installation of software of any sort on the server and requires no server side disk space.[02.04] What server side operating systems does DataBee support?All of them actually. DataBee can be used to extract and load data onto any server that will run an Oracle database. DataBee is server neutral, it runs on a PC and connects via the Oracle supplied SQL*Net package to the source and destination databases. The server side operating system is largely irrelevant to it.[02.05] What Oracle database versions does DataBee support?DataBee supports all versions of 7, 8, 9 and 10 of the Oracle database servers. In fact, you can even extract from versions 7 or 8 of an Oracle database and load onto a version 10 Oracle database.[02.06] What version of DataBee am I currently using?Every copy of DataBee contains version information in the "About DataBee" dialog box. To see this dialog box start any DataBee application and click on the Bee logo in the upper right hand corner.[02.07] How can I see or edit my License key?To view the License Key dialog box press the Registration Key button in the "About DataBee" dialog box. See the FAQ question [02.06] for information on the "About DataBee" dialog box.[02.08] How do I upgrade DataBee?Just get the latest version from the web site (http://www.DataBee.com) and re-install. You do not need to uninstall first and you can skip as many versions as you wish. Your saved settings and license keys will be preserved.
[03.01] How big can a DataBee Extracted Set File get? 'How do I' and 'Can I' Type Questions
Extracted Set Files hold the identifiers of the rows in the source database that will need to be loaded onto the destination in order to create the subset database. There are a number of compression and storage efficiencies in place so that you will find that while an Extracted Set File will initially grow very quickly, the growth rate will slow down as more rows are added. Having said that, Extracted Set Files of 80Mb or more are common for large extracts - it all depends on the number of rows extracted.[03.02] Can I share my Extracted Set File with other users?Yes, however, be aware that the Extracted Set Files are viable only as long as the source database has not changed. This is because DataBee does not store the row contents in the Extracted Set File - just the row identifiers. If the source database changes after the Extracted Set File is taken, but before the load is performed, then the rows in the Extracted Set File may be out of synchronization with the rows that are currently required. A better idea is to share your extraction set (the extraction rules) with other users and let them use it to perform their own extracts.[03.03] Can I put my extraction directory on a Network Drive?Yes, but it is a bad idea. Network drives are much slower than local drives and DataBee will take longer to read and write to an Extracted Set File on a network drive. A better idea is to place it on your local drive during the extract and then copy it over once the extract is complete.[03.04] Can a table have more than one WHERE Clause rule?Absolutely, place as many WHERE Clause rules on a table as you need to extract the rows you wish. You do not need to worry about extracting duplicate rows either - DataBee will automatically take care of removing them. See the About Rules page for more information.[03.05] Can a table have more than one rule pointing out of it?Yes, DataBee is specifically designed for this. See the About Rules page for more information.[03.06] Can a table have more than one rule pointing into it?Yes, DataBee is specifically designed for this. See the About Rules page for more information.[03.07] Can a table have a rule pointing at itself?Yes, DataBee easily handles self-referential rules.[03.08] Can a table have more than one rule pointing at another table based on different join conditions?Yes, DataBee will take each rule in turn and make sure the Child table has the rows it needs in order to satisfy both rules.[03.09] Can a table have more than one rule pointing at another table based on the same join condition?No, earlier versions of DataBee permitted this - but this redundancy is now caught and flagged as an error. If the duplicate rule exists in an older extraction set it will still be processed as normal.[03.10] If table A is the Parent of table B in one rule can it be the Child of table B in another rule?Yes, DataBee will take each rule in turn and make sure the Child table has the rows it needs in order to satisfy both rules.[03.11] Why does the Set Designer sometimes not know what columns a table has?This can happen if you do not have any schema information associated with an extraction set. Details of the tables and columns in a schema are saved in a .schema file while the set is not loaded in the Set Designer application. If the .schema file is absent or incomplete, the Set Designer will know from the extraction set rules some of the names of the tables in the schema. However, it will not know the column information or know anything about any of the tables in the schema that are not involved in rules. Use the Get Schema Information button on the Set Designer Rule Tools tab to download the schema information again.[03.12] The rest of the rules extract really fast. Why is one so slow?[03.13] How do I change the directory in which the log files are placed?
- Are both tables in the rule analyzed? Odds are one or both is not and Oracle is simply putting together an inefficient execution plan. Find the names of both tables by using the "Get Rule Information" button in the Rule Statistics tab and try the extract again. In fact it is a good idea to analyze the entire schema before doing a DataBee Extract. If you need to do so you can also apply an Oracle SQL hint to the statement.
- Are the join columns used in the rule indexed in the child table? If the join columns in the target table are not indexed the only option Oracle will have is to do a full table scan. This will be slow if there are a lot of rows in the target table. You can easily see which rules have unindexed join columns using the Set Designer Unindexed Rules Detector Tool. On v03.00 (and greater) of the DataBee Set Designer a prominently displayed icon turns green or red when viewing the rule columns to indicate the indexed status of the columns used in the rule.
- For reasons that remain a mystery to us - Oracle sometimes chooses a really bad plan for some rules. We cannot see any reason - the tables and indexes are analyzed, the optimizer mode is COST and the target columns are indexed - and yet plans requiring full table scans are generated. In these circumstances adding a "/*+ RULE*/" hint to the DataBee rule (it is simple to do) always sorts things out and the extract gets a massive performance increase.
By default, the log files are placed in a directory named DataBee Log Files which is located just below the DataBee installation directory. The log directory must be common to all DataBee applications. The log file directory for all DataBee applications can be changed using the Log File Directory option located on the Set Designer Rule Tools tab.[03.14] Do I have to have a plan in order to create an extraction set?Plans are recommended but not strictly necessary. To create an extraction set without a plan just use the Set Designer New Set Wizard as usual but say "No" when asked if you have a plan. You will be asked for more information as you progress through the wizard panels but the process is quite intuitive.[03.15] My database has no foreign keys, can DataBee still extract a subset?The DataBee Set Designer application will automatically generate extraction rules to enforce foreign key constraints. Thus you are guaranteed that the constraints will enable in the subset database.[03.16] Can I create a subset on a earlier version of Oracle than the source database?Foreign keys are not necessary in order to build an extraction set. If they are present in the schema then you must use them in the extraction set. If foreign key constraints are not present, then DataBee has lots of discovery tools which enable you to generate your own rules.
In theory, no, in practice, yes. All DataBee really requires is that the table structures be identical. For example, it is possible for tables on a version 9i database to contain structures and objects not physically possible on a version 7 database. For this reason we say you can create a subset of an Oracle version 7 database on a version 9 platform but not a version 9 subset on a version 7 platform.[03.17] Can I create a referentially correct set of data from tables distributed between more than one schema?In reality, if your source database structure doesn't contain any features not implementable on the earlier Oracle version then you will probably find that the subset creates just fine whichever way around you do it.
Chapter 2 of the DataBee Quick Start Guide contains a more in-depth discussion of this topic
Yes, version 03.00 (and greater) of the DataBee software handles this transparently.[03.18] What do the letters FP mean in the extracted row count columns?It means the table is to be loaded via the Full-Pull method and that you have set the options in the Set Designer application to indicate that this table should not bother extracting its rows. More information can be found on this help page.[03.19] Can DataBee extract and load tables containing user defined column types or LONGS?Yes it can, in fact DataBee handles such tables transparently. For tables containing columns of a user defined type (objects, varrays, nested tables etc or LONG columns) DataBee invokes low level internal routines that completely bypass Oracles normal transfer mechanisms. This means that you do not get ORA-00997, ORA-22804 or ORA-00932 errors while transferring data.
[Net 2000 Ltd. Home][DataBee Home][DataBee Manual]
