Ext3 commits writes to disk within approximately 5 seconds - Ext4 can take from 40-150 seconds. In addition, if a system is using Ext3 and crashes before the commit takes place you will still have the previous contents of a file where under Ext4 the file will be empty. Theodore Tso feels that this is a failure at the application level and that the file system is behaving as designed and as specified by the POSIX spec (which apparently does not specify what is supposed to happen when a system is not shut down cleanly). His solution to the issue is to suggest proper use of fsync() and lists various scenarios/examples in post 54 of the bug report (linked above). In addition he wrote a patch that recognize the rename() situation mentioned in his post 54 yet retains the normal Ext4 behaviors and performance in the majority of cases. Also a more "proper" solution has been provided which allows the behavior of Ext3 to be retained under Ext4 by mounting it with alloc_on_commit.
A somewhat related topic is the use of on-board caching by hard drives. This behavior can be modified on most drives by using hdparm.
In this article Barry Mavin explains step by step how to setup a Linux HA (High Availability) cluster for the running of Recital applications on Redhat/Centos 5.3 although the general configuration should work for other linux versions with a few minor changes.
Binary distributions for Unison can be found here.
The user manual can be found here.
In this article Barry Mavin, CEO and Chief Software Architect for Recital, details on how to use the Client Drivers provided with the Recital Database Server to work with local or remote server-side JDBC data sources.
Overview
The Recital Universal .NET Data Provider provides connectivity to the Recital Database Server running on any supported platform (Windows, Linux, Unix, OpenVMS) using the RecitalConnection object.
The Recital Universal JDBC Driver provides the same functionality for java applications.
The Recital Universal ODBC Driver provides the same functionality for applications that use ODBC.
Each of the above Client Drivers use a connection string to describe connections parameters.
The basic format of a connection string consists of a series of keyword/value pairs separated by semicolons. The equals sign (=) connects each keyword and its value.
The following table lists the valid names for keyword/values.
| Name | Default | Description |
|---|---|---|
|
Data Source |
The name or network address of the instance of the Recital Database Server which to connect to. | |
| Directory | The target directory on the remote server where data to be accessed resides. This is ignored when a Database is specified. | |
|
Encrypt |
false | When true, DES3 encryption is used for all data sent between the client and server. |
| Initial Catalog -or- Database |
The name of the database on the remote server. | |
| Password -or- Pwd |
The password used to authenticate access to the remote server. | |
| User ID | The user name used to authenticate access to the remote server. | |
|
Connection Pooling |
false | Enable connection pooling to the server. This provides for one connection to be shared. |
| Logging | false | Provides for the ability to log all server requests for debugging purposes |
| Rowid | true | When Rowid is true (the default) a column will be post-fixed to each SELECT query that is a unique row identifier. This is used to provide optimised UPDATE and DELETE operations. If you use the RecitalSqlGrid, RecitalSqlForm, or RecitalSqlGridForm components then this column is not visible but is used to handle updates to the underlying data source. |
| Logfile | The name of the logfile for logging | |
| Gateway |
Opens an SQL gateway(Connection) to a foreign SQL data source on
the remote server.
servertype@nodename:username/password-database e.g. oracle@nodename:username/password-database mysql@nodename:username/password-database postgresql@nodename:username/password-database -or- odbc:odbc_data_source_name_on_server oledb:oledb_connection_string_on_server jdbc:jdbc_driver_path_on_server;jdbc:Recital:args |
To connect to a server-side JDBC data source, you ue the gateway=value key/value pair in the following way.
gateway=jdbc:jdbc_driver_path_on_server;jdbc:Recital:args
You can find examples of connection strings for most ODBC and OLE DB data sources by clicking here.
Example in C# using the Recital Universal .NET Data Provider:
////////////////////////////////////////////////////////////////////////
// include the references below
using System.Data;
using Recital.Data;
////////////////////////////////////////////////////////////////////////
// The following code example creates an instance of a DataAdapter that
// uses a Connection to the Recital Database Server, and a gateway to
// Recital Southwind database. It then populates a DataTable
// in a DataSet with the list of customers via the JDBC driver.
// The SQL statement and Connection arguments passed to the DataAdapter
// constructor are used to create the SelectCommand property of the
// DataAdapter.
public DataSet SelectCustomers()
{
string gateway = "jdbc:/usr/java/lib/RecitalJDBC/Recital/sql/RecitalDriver;"+
"jdbc:Recital:Data Source=localhost;database=southwind";
RecitalConnection swindConn = new
RecitalConnection("Data Source=localhost;gateway=\""+gateway+"\");
RecitalCommand selectCMD = new
RecitalCommand("SELECT CustomerID, CompanyName FROM Customers", swindConn);
selectCMD.CommandTimeout = 30;
RecitalDataAdapter custDA = new RecitalDataAdapter();
custDA.SelectCommand = selectCMD;
swindConn.Open();
DataSet custDS = new DataSet();
custDA.Fill(custDS, "Customers");
swindConn.Close();
return custDS;
}
Example in Java using the Recital Universal JDBC Driver:
//////////////////////////////////////////////////////////////////////// // standard imports required by the JDBC driver import java.sql.*; import java.io.*; import java.net.URL; import java.math.BigDecimal; import Recital.sql.*; ////////////////////////////////////////////////////////////////////////
// The following code example creates a Connection to the Recital // Database Server, and a gateway to the Recital Southwind database. // It then retrieves all the customers via the JDBC driver. public void SelectCustomers() { // setup the Connection URL for JDBC String gateway = "jdbc:/usr/java/lib/RecitalJDBC/Recital/sql/RecitalDriver;"+ "jdbc:Recital:Data Source=localhost;database=southwind"; String url = "jdbc:Recital:Data Source=localhost;gateway=\""+gateway+"\";
// load the Recital Universal JDBC Driver new RecitalDriver(); // create the connection Connection con = DriverManager.getConnection(url); // create the statement Statement stmt = con.createStatement(); // perform the SQL query ResultSet rs = stmt.executeQuery("SELECT CustomerID, CompanyName FROM Customers"); // fetch the data while (rs.next()) { String CompanyID = rs.getString("CustomerID"); String CompanyName = rs.getString("CompanyName"); // do something with the data... } // Release the statement stmt.close(); // Disconnect from the server con.close(); }
In this article Barry Mavin, CEO and Chief Software Architect for Recital, details how to work with Triggers in the Recital Database Server.
Overview
A trigger is a special kind of stored procedure that runs when you modify data in a specified table using one or more of the data modification operations: UPDATE, INSERT, or DELETE.
Triggers can query other tables and can include complex SQL statements. They are primarily useful for enforcing complex business rules or requirements. For example, you can control whether to allow a new order to be inserted based on a customer's current account status.
Triggers are also useful for enforcing referential and data integrity.
Triggers can be used with any data source that is handled natively by the Recital Database Engine. This includes Recital, FoxPro, FoxBASE, Clipper, dBase, CISAM, and RMS data,
Creating and Editing Triggers
To create a new Trigger, right-click the Procedures node in the Databases tree of the Project Explorer and choose Create. To modify an existing Trigger select the Trigger in the Databases Tree in the Project Explorer by double-clicking on it, or select Modify from the context menu. By convertion we recommend that you name your Stored Procedures beginning with "sp_xxx_", user-defined functions with "f_xxx_", and Triggers with "dt_xxx_", where xxx is the name of the table that they are associated with.
Associating Triggers with a Table
Once you have written your Triggers as detailed above you can associate them with the operations performed on a Table by selecting the Table tab.
The Tables tab allows you to select a Trigger procedure by clicking on the small button at the right of the Text field.
Types of Triggers
As can be seen from the Tables tab detailed below, The Recital Database Server handles 6 distinct types of Triggers.
Open Trigger
The Open Trigger is called after is a table is opened but before any operations are performed on it. You can use this trigger to record a log of table usage or provide a programmable means of checing security. If the Trigger procedure returns .F. (false), then the table is not opened. You can use a TRY...CATCH block around the associated command to inform the user.
Close Trigger
The Close Trigger is called just prior to a table being closed. In this trigger you may find it useful to get transaction counts by using the IOSTATS() built-in 4GL function, and record these values in a transaction log.
Update Trigger
The Update Trigger is called prior to a record update operation being performed. You can use this trigger to perform complex application or data specific validation. If the Trigger procedure returns .F. (false), then the record is not updated. You can use inform the user from within the Trigger procedure the reason that the data cannot be updated.
Delete Trigger
The Delete Trigger is called prior to a record delete operation being performed. You can use this trigger to perform complex application or data specific validation such as cross-table lookups e.g. attempting to delete a customer recortd when there are still open orders for that specific customer. If the Trigger procedure returns .F. (false), then the record is not deleted.
Insert Trigger
The Insert Trigger is called prior to a record insert (append) operation being performed. You can use this trigger to perform such tasks as setting up default values of columns within the record. If the Trigger procedure returns .F. (false), then the record is not inserted.
Rollback Trigger
The RollbackTrigger is called prior to a rollback operation being performed from within a form. If the Trigger procedure returns .F. (false), then the record is not rolled back to its original state.
Testing the Trigger
To test run the Trigger, select the Trigger in the Databases Tree in the Project Explorer by double-clicking on it. Once the Database Administrator is displayed, click the Run button to run the Trigger.
The REQUIRE() statement includes and executes the contents of the specified file at the current program execution level.
When a file is included, the code it contains inherits the variable scope of the line on which the include occurs. Any variables, procedures, functions or classes declared in the included file will be available at the current program execution level.
The REQUIRE_ONCE() statement is identical to the REQUIRE() statement except that Recital will check to see if the file as already been included and if so ignore the command.
The full syntax is:
REQUIRE( expC ) REQUIRE_ONCE( expC ) e.g. REQUIRE_ONCE( "myapp/myglobals.prg" )
http://kbala.com/ie-9-supports-corner-radius/
A quick tip for optimizing TCP performance on linux.
edit /etc/sysctl.conf add the lines:
If using gigabit ethernet:
net.ipv4.tcp_mem= 98304 131072 196608
net.ipv4.tcp_window_scaling=1
net.core.wmem_default = 65536
net.core.rmem_default = 65536
net.core.wmem_max=8388608
To reload these use:
# sysctl -p
If using infiniband:
net.ipv4.tcp_window_scaling=1
net.ipv4.tcp_timestamps=0
net.ipv4.tcp_sack=0
net.ipv4.tcp_rmem=10000000 10000000 10000000
net.ipv4.tcp_wmem=10000000 10000000 10000000
net.ipv4.tcp_mem=10000000 10000000 10000000
net.core.rmem_max=524287
net.core.wmem_max=524287
net.core.rmem_default=524287
net.core.wmem_default=524287
net.core.optmem_max=524287
net.core.netdev_max_backlog=300000
RTOS()
Syntax
RTOS( [ <workarea> ] )Description
The RTOS() function returns all the fields in the current row as a string. The string will begin with the unique row identifier and then the deleted flag, followed by the data in the record. An optional workarea can be specified, otherwise the current workarea will be usedExample
use backup in 0
use accounts in 0
nrecs=reccount()
for i = 1 to nrecs
if rtos(accounts) != rtos(backup)
debug("record "+recno()+" don't match")
endif
next
This article discusses the features in Recital that allow data to be imported and exported between platforms in Microsoft® ADO XML Format.
Overview
Extensible Markup Language, XML, is widely regarded as a lingua franca for the interchange of data. XML's text-based, platform-independent format and its integration of data and the schema to define and describe that data, make it the ideal import/export medium. Recital software provides the functionality to output the data from Recital - and other supported table formats such as FoxPro and FoxBASE - into XML file format and to import XML data into those tables' formats. Such import/export operations provide the means to exchange data with third-party applications and can also facilitate the transfer of data between Recital installations on binary-incompatible platforms.
The features examined in this article are available in Recital Terminal Developer and in the Recital Mirage and Recital Database Servers on all Recital supported platforms. Both the Recital/4GL and Recital/SQL provide XML import and export capabilities. The XML files discussed are in Microsoft® ADO XML format.
Microsoft® ActiveX® Data Objects XML Format
The ADO XML format is primarily designed for ADO Recordset persistence and ADO XML files created by Recital can be used in this way and loaded directly into ADO Recordsets. The format can, though also be used for more generic data transfer. An ADO XML file is self-contained, consisting of two sections: a schema section followed by a data section. The schema conforms to the W3C XML-Data specification and defines the data structure.
For additional information on the Microsoft® ActiveX® Data Objects XML Format, please see Appendix 1.
NOTE: The Recital XMLFORMAT setting should always be in its default setting of ADO for ADO XML Format operations.
set xmlformat to ADO
SQL
Recital/SQL offers the ability to export data into XML files using the SELECT and FETCH statements and import from XML using the CREATE TABLE and INSERT statements.
SQL: Exporting
The SELECT...SAVE AS XML statement allows the complete result set from a SELECT statement to be saved as an XML file. This could be a complete table:
open database southwind SELECT * from orders SAVE AS XML orders.xml
or a more complex multi-table query:
open database southwind
SELECT orders.orderid, orders.customerid,;
employees.employeeid, employees.lastname, employees.firstname,;
orders.orderdate, orders.freight, orders.requireddate,;
orders.shippeddate, orders.shipvia, orders.shipname,;
orders.shipaddress, orders.shipcity,;
orders.shipregion, orders.shippostalcode, orders.shipcountry,;
customers.companyname, customers.address, customers.city,;
customers.region, customers.postalcode, customers.country;
FROM orders INNER JOIN customers;
ON customers.customerid = orders.customerid,;
orders INNER JOIN employees;
ON orders.employeeid = employees.employeeid;
SAVE AS XML orderinfo
The resulting XML file can then be further processed within the same or a different Recital environment or transferred to a third party product.
<x-ml xmlns:z="#RowsetSchema" xmlns:rs="urn:schemas-microsoft-com:rowset" xmlns:dt="uuid:C2F41010-65B3-11d1-A29F-00AA00C14882" xmlns:s="uuid:BDC6E3F0-6DA3-11d1-A2A3- 00AA00C14882"> <s:schema id="RowsetSchema"> <s:elementtype rs:updatable="true" content="eltOnly" name="row"> <s:attributetype rs:basecolumn="orderid" rs:basetable="orders.dbf" rs:write="true" rs:nullable="true" rs:number="1" name="orderid"> <s:datatype rs:fixedlength="true" rs:precision="14" rs:scale="0" dt:maxlength="10" rs:dbtype="numeric" dt:type="number"> </s:datatype></s:attributetype> <s:attributetype rs:basecolumn="customerid" rs:basetable="orders.dbf" rs:write="true" rs:nullable="true" rs:number="2" name="customerid"> <s:datatype rs:fixedlength="true" dt:maxlength="5" rs:dbtype="str" dt:type="string"> </s:datatype></s:attributetype> <s:attributetype rs:basecolumn="employeeid" rs:basetable="orders.dbf" rs:write="true" rs:nullable="false" rs:number="3" name="employeeid"> <s:datatype rs:fixedlength="true" rs:precision="20" rs:scale="0" dt:maxlength="10" rs:dbtype="numeric" dt:type="number"> </s:datatype></s:attributetype> <s:attributetype rs:basecolumn="lastname" rs:basetable="orders.dbf" rs:write="true" rs:nullable="false" rs:number="4" name="lastname"> <s:datatype rs:fixedlength="true" dt:maxlength="20" rs:dbtype="str" dt:type="string"> </s:datatype></s:attributetype> <s:attributetype rs:basecolumn="firstname" rs:basetable="orders.dbf" rs:write="true" rs:nullable="false" rs:number="5" name="firstname"> <s:datatype rs:fixedlength="true" dt:maxlength="10" rs:dbtype="str" dt:type="string"> </s:datatype></s:attributetype> <s:attributetype rs:basecolumn="orderdate" rs:basetable="orders.dbf" rs:write="true" rs:nullable="true" rs:number="6" name="orderdate"> <s:datatype rs:fixedlength="true" dt:maxlength="10" rs:dbtype="Date" dt:type="Date"> </s:datatype></s:attributetype> <s:attributetype name="freight" ...
Click image to display full size
Fig 1: Microsoft® Office Excel 2003: orderinfo.xml.
For data accessed through a Recital Database Gateway, such as Oracle, MySQL or PostgreSQL, the FETCH command can be used to save a cursor results set into an XML file:
// Connect to MySQL Database 'mydata' via Recital Database Gateway
nStatHand=SQLSTRINGCONNECT("mys@mysql1:user1/pass1-mydata",.T.)
if nStatHand < 1
dialog box [Could not connect]
else
DECLARE cursor1 CURSOR FOR;
SELECT account_no, last_name, first_name FROM example
OPEN cursor1
FETCH cursor1 INTO XML exa1.xml
SQLDISCONNECT(nStatHand)
endif
SQL: Importing
The CREATE TABLE statement allows a new table to be created based on the structure defined in an XML file. The data from the XML file can optionally be loaded into this new table if the LOAD keyword is included. For example, a new 'orderinfo' table can be created and populated with data from the orderinfo.xml file created by the SELECT...SAVE AS XML statement shown earlier:
open database southwind
SELECT orders.orderid, orders.customerid,;
employees.employeeid, employees.lastname, employees.firstname,;
orders.orderdate, orders.freight, orders.requireddate,;
orders.shippeddate, orders.shipvia, orders.shipname,;
orders.shipaddress, orders.shipcity,;
orders.shipregion, orders.shippostalcode, orders.shipcountry,;
customers.companyname, customers.address, customers.city,;
customers.region, customers.postalcode, customers.country;
FROM orders INNER JOIN customers;
ON customers.customerid = orders.customerid,;
orders INNER JOIN employees;
ON orders.employeeid = employees.employeeid;
SAVE AS XML orderinfo
CREATE TABLE orderinfo FROM XML orderinfo LOAD
The INSERT statement can be used to load data when the table structure already exists. Taking our earlier orderinfo.xml file again, the data can be loaded using INSERT:
open database southwind;
SELECT orders.orderid, orders.customerid,;
employees.employeeid, employees.lastname, employees.firstname,;
orders.orderdate, orders.freight, orders.requireddate,;
orders.shippeddate, orders.shipvia, orders.shipname,;
orders.shipaddress, orders.shipcity,;;
orders.shipregion, orders.shippostalcode, orders.shipcountry,;
customers.companyname, customers.address, customers.city,;
customers.region, customers.postalcode, customers.country;
FROM orders INNER JOIN customers;
ON customers.customerid = orders.customerid,;
orders INNER JOIN employees;
ON orders.employeeid = employees.employeeid;
SAVE AS XML orderinfo
CREATE TABLE orderinfo FROM XML orderinfo
INSERT INTO orderinfo FROM XML orderinfo
The examples above show the export and import in a single piece of code. To transfer data between binary-incompatible platforms, the export phase using SELECT...SAVE AS XML would be carried out on the source platform, the resulting XML file would be transferred to the target platform, then the import phase using CREATE TABLE...LOAD or CREATE TABLE + INSERT would be run on the target platform.
Recital/4GL
The Recital/4GL offers the ability to export data into XML files using the COPY TO ... TYPE XML command and import from XML using the XMLFIRST() and XMLNEXT() functions.
Recital/4GL: Exporting
The COPY TO command can be used to export data from Recital and other natively supported tables out to a wide range of formats. This includes exporting to an XML file. The '.xml' file extension is added automatically. The COPY TO command can be used to export an entire table:
open database southwind use orders copy to orders type xml
or, using the FIELDS clause and the FOR or WHILE clauses, restrict the field list and export only those records which match a particular condition:
open database southwind use orders copy to orders type xml fields orderid for year(orderdate) = 1996
Only the orderid field from those records which match the condition is exported:
<x-ml xmlns:z="#RowsetSchema" xmlns:rs="urn:schemas-microsoft-com:rowset" xmlns:dt="uuid:C2F41010-65B3-11d1-A29F-00AA00C14882" xmlns:s="uuid:BDC6E3F0-6DA3-11d1-A2A3- 00AA00C14882"> <s:schema id="RowsetSchema"> <s:elementtype rs:updatable="true" content="eltOnly" name="row"> <s:attributetype rs:basecolumn="ORDERID" rs:basetable="ORDERS" rs:write="true" rs:nullable="true" rs:number="1" name="ORDERID"> <s:datatype rs:fixedlength="true" rs:precision="10" rs:scale="0" dt:maxlength="10" rs:dbtype="numeric" dt:type="number"> </s:datatype></s:attributetype> </s:elementtype> </s:schema> <rs:data> <z:row orderid="10248"> <z:row orderid="10249"> <z:row orderid="10250"> <z:row ...
Recital/4GL: Importing
Data from an XML file can be extracted one record at a time using the XMLFIRST() and XMLNEXT() functions. XMLFIRST() reads the first record from an XML file and loads information from the file into a series of memory variables and arrays. The record data is loaded into a one-dimensional array which is created automatically. Each element in the array contains the data for its corresponding field in string format. The field names are loaded into another automatically-created array. The XMLNEXT() function works in a similar way to deal with all the subsequent records in the XML file. The XMLCOUNT() function can be used, as in the example below, to determine how many data records the XML file has.
The Recital/4GL includes a vast range of functions for manipulation and conversion of arrays and their individual elements. In the example program below, the XMLFIRST() and XMLNEXT() functions are used to sequentially extract each record from an XML file, whose name is passed to the program as a parameter. Once loaded into an array, the data is converted to the correct Recital data type then appended into a table. The table name is also passed as a parameter.
procedure replaceit
append blank
for i = 1 to numfields
if type(field(i)) = "N"
replace &(field(i)) with val(data[&i])
elseif type(field(i)) = "D"
replace &(field(i)) with stod(data[&i])
elseif type(field(i)) = "T"
replace &(field(i)) with ctot(data[&i])
elseif type(field(i)) = "L"
replace &(field(i)) with iif(data[&i]="T",.T.,.F.)
elseif type(field(i)) = "Y"
replace &(field(i)) with val(data[&i])
else
replace &(field(i)) with data[&i]
endif
next
return
procedure starthere
parameters cTable, cFile
numfields=xmlfirst(cFile,targ,trans,where,fldnames,data)
if numfields < 1
dialog box [No records in XML file]
else
use &cTable
replaceit()
endif
numrecs = xmlcount(cFile)
if numrecs > 1
numleft = numrecs -1
for i = 1 to numleft
xmlnext(trans,where,fldnames,data)
replaceit()
next
endif
return
Alternative Import/Export Methods
Other features exist in Recital to facilitate the import and export of data:
RDDs
The RDDs, Replaceable Database Drivers, are available on Windows, Linux and all supported 32-bit UNIX platforms. They allow for the use and creation of database tables and indexes in FoxPro, dBase and Clipper formats. The file format is the same across all the platforms that support the RDDs, allowing the tables and indexes to be transferred as required. The formats are also supported by a wide range of third-party products as well as their originating database systems. For more information on the RDDs, please see the online documentation on Xbase migration and the SET FILETYPE command.
BUILD/INSTALL
These are Recital/4GL commands for the export (BUILD) and import (INSTALL) of Recital tables and their associated memo, dictionary and multiple index files in ASCII format to allow them to be transferred across binary incompatible platforms. For more information, please see the online documentation on Recital/4GL commands.
COPY Commands
The COPY TO, COPY STRUCTURE, COPY STRUCTURE EXTENDED and CREATE FROM commands can all be used to enable data to be transferred between different formats and different platforms. For more information, please see the online documentation on Recital/4GL commands.
Appendix 1: Microsoft® ActiveX® Data Objects XML Format
For detailed information on the Microsoft® ActiveX® Data Objects XML Format, please consult the following Microsoft documentation:
|
Link |