diff options
Diffstat (limited to 'ACE/TAO/docs/releasenotes/index.html')
| -rw-r--r-- | ACE/TAO/docs/releasenotes/index.html | 2560 |
1 files changed, 2560 insertions, 0 deletions
diff --git a/ACE/TAO/docs/releasenotes/index.html b/ACE/TAO/docs/releasenotes/index.html new file mode 100644 index 00000000000..ee93e683c2a --- /dev/null +++ b/ACE/TAO/docs/releasenotes/index.html @@ -0,0 +1,2560 @@ +<!doctype html public "-//w3c//dtd html 4.0 transitional//en"> +<html> +<head> + <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1"> + <meta name="GENERATOR" content="Mozilla/4.79 [en] (Windows NT 5.0; U) [Netscape]"> + <title>TAO Release Information and TODO List</title> +</head> +<body text="#000000" bgcolor="#FFFFFF"> +<!-- $Id$ --> +<center> +<hr></center> + +<center> +<h3> +Release Information for The ACE ORB (TAO)</h3></center> +This document contains information on the following topics related to the +<a href="../../VERSION">current +release</a> of <a href="http://www.dre.vanderbilt.edu/TAO">TAO</a>: + +<table BORDER=0 CELLSPACING=0 CELLPADDING=10 > +<tr> +<td VALIGN=TOP> +<b>ACE Related</b> +<br> +<ul> +<li> +<a href="#ace">ORB-related ACE Changes</a></li> +</ul> +</td> +<td VALIGN=TOP> + +<b>Misc Entries</b> +<br> +<ul> + +<li> +<a href="#CORBA-conformance">CORBA Standards Conformance</a></li> + + +<li> +<a href="#dove">The DOVE Demo</a></li> + +<li> +<a href="TODO.html">TODO list</a></li> +</ul> +</td> + +</table> + +<table BORDER=0 CELLSPACING=0 CELLPADDING=10 > +<tr> +<td VALIGN=TOP> +<b>ORB Related</b> +<br> +<ul> +<li> +<a href="amh.html">Asynchronous Method Handling (AMH)</a></li> + +<li> +<a href="#ami">Asynchronous Method Invocation (AMI)</a></li> + +<li> +<a href="#csd">Custom Servant Dispatching (CSD)</a></li> + +<li> +<a href="../dynany/index.html">Dynamic Any</a></li> + +<li> +<a href="#leader">Global Resources and Leader-Follower Model</a></li> + +<li> +<a href="#idl">IDL Compiler</a></li> + +<li> +<a href="../implrepo/index.html">Implementation Repository</a></li> + +<li> +<a href="../interfacerepo/index.html">Interface Repository</a></li> + +<li> +<a href="#localobject">Local Interfaces</a></li> + +<li> +<a href="#locate">Locate requests</a></li> +<li> +<a href="#forwarding">Location Forwarding</a></li> + + + +<li> +<a href="#miop">Multicast InterORB Protocol (MIOP)</a></li> + +<li> +<a href="orbcore.html">ORB Core</a></li> + +<li> +<a href="OBV.html">Object-by-Value</a></li> + +<li> +<a href="../poa_migration.html">POA Migration Notes</a></li> + + +<li> +<a href="#pp">Pluggable Protocols</a></li> + +<li> +<a href="#interceptor">Portable Interceptors</a></li> + +<li> +<a href="#poa">Portable Object Adapter (POA)</a></li> + + +<li> +<a href="../rtcorba/index.html">Real-Time CORBA</a></li> + +<li> +<a href="#sciop">SCIOP Support in TAO</a></li> + +<li> +<a href="#ipv6">IPv6 Support in TAO</a></li> + +<li> +<a href="../Smart_Proxies.html">Smart Proxies</a></li> +</ul> +<td VALIGN=TOP> +<b>CORBA Services Related</b> +<br> +<ul> +<li> +<a href="#av">Audio/Video Streaming Service</a></li> +<li> +<a href="#cservices">Concurrency Service</a></li> +<li> Event Service +<ul> +<li> +<a href="#ec">CORBA Event Service</a></li> +<li> +<a href="ftrt_ec.html">Fault_Tolerant/Real_Time Event Service</a></li> +<li> +<a href="ec.html">Real-time Event Service</a></li> +</ul> + +<li> +<a href="#fault_tolerance">Fault Tolerant Services </a></li> + +<li> +<a href="#loadbalancer">Load Balancer Service</a></li> + + +<li> +<a href="#nservices">Naming Service </a></li> +<li> +Notification Service +<ul> +<li> +<a href="#notify">Notification Service</a></li> +<li> +<a href="#rt_notify">Real-Time Notification Service</a></li> +</ul> +<li> +<a href="PSS.html">Persistent State Service</a></li> +<li> +<a href="#pservices">Property Service</a></li> +<li>Scheduling Service +<ul> +<li> +<a href="#scheduling">Kokyu Scheduling Service</a></li> +<li> +<a href="rtc10_sched.html">RTCORBA 1.0 Scheduling Service</a></li> +</ul> + +<li> +<a href="#security">Security Service</a></li> + +<li> +<a href="#log">Telecom Log Service</a></li> +<li> +<a href="#ts">Time Service</a></li> +<li> +<a href="#tservices">Trading Service</a></li> + + + +</ul> +</td> + +<td VALIGN=TOP> +<ul> +<b>CCM Related</b> +<br><BR> +<li> +<a href="../../../CIAO/docs/releasenotes/index.html">CORBA Component Model (CIAO)</a></li> +<li> +<a href="../../../CIAO/docs/releasenotes/dance.html">Deployment and + Configuration Engine (DAnCE)</a></li> +<li> +<a href="../../../CIAO/CIDLC/README.html">Component Implementation +Definition Language Compiler (CIDLC)</a></li> +<li> +<a href="../../../CIAO/docs/static_ciao_contents.html">Static Configuration + Support for Real-Time Platforms</a></li> +</ul> +</td> + +<td VALIGN=TOP> +</td> +</tr> +</table> + + +<table BORDER=0 CELLSPACING=0 CELLPADDING=10 > +<tr> + + +</table> + + +A complete list of all modifications to TAO is available in the <a href="../../ChangeLog">ChangeLog</a>. +<p> +<hr> +<h3> +<a NAME="ace"></a>ACE Wrappers</h3> +Current status: (As of July 07, 2003.) +<ul> +<li> +The following changes are related to improving memory management of Event +Handlers when they interact with Reactors (and Timer Queues). When a handler +is registered with the Reactor, the Reactor increments the reference count +on the handler. The Reactor also increments this reference count when making +upcalls on the handler. The reference count is decremented when an upcall +completes or when the handler is removed from the Reactor.</li> + +<p>This mechanism is similar to what happens between POAs and Servants +and it allows for the safe deletion of handlers. This mechanism is particularly +need for multi-threaded applications that can have multiple threads executing +upcalls on a handler that needs to be shutdown in a safe manner. The following +illustrates an example of how this mechanism works: +<ul> +<li> +When a handler is created, it reference count is one.</li> + +<li> +After the handler is registered with the Reactor, it reference count becomes +two.</li> + +<li> +At this point, the handler creator can let go of the handler reference, +bringing down the reference count to one.</li> + +<li> +For each thread executing upcalls on the handler, the Reactor increments +the reference count by one. So if three threads were simultaneously making +upcalls on a handler, the reference count would be four.</li> + +<li> +Assuming an external event (and thread) decides to close the handler. It +simply removes the handler from the Reactor. This decreases the reference +count to three.</li> + +<li> +As each thread completes their upcall, the reference reduces.</li> + +<li> +Once the final thread exits the upcall, the reference reaches zero, and +the handler is finally deleted.</li> +</ul> +This mechanism ensures that the handler is not deleted until the final +upcall thread exits the handler. +<p>Reference counting on handlers is optional and is disabled by default. +To enable reference counting on a handler, reset its reference counting +policy to "ENABLED". +<p>To facilitate reference counting of handlers, an ACE_Event_Handler_var +class was added. This class is akin to the PortableServer::ServantBase_var +class. +<p>Similar reference counting related changes were made to the Timer Queues +so that handlers can be used in a thread safe manner with the queues. +<p>The Connector implementation was completely revised to utilize the new +memory management mechanisms and remove existing concurrency bugs. +<p>Several new examples/tests were added: Reference_Counted_Event_Handler_Test, +MT_Reference_Counted_Event_Handler_Test, MT_Reference_Counted_Notify_Test, +Timer_Queue_Reference_Counting_Test, NonBlocking_Conn_Test, Reactor_Registration_Test, +WFMO_Reactor_Test, Timer_Cancellation_Test</ul> + +<hr> +<h3> +<a NAME="idl"></a>IDL Compiler</h3> +Point of contact: <a href="mailto:j.parsons@vanderbilt.edu">Jeff Parsons</a> +<p>Current status: (As of November 13, 2006.) +<ul> +<li> +Generated code closely follows the C++ Mapping specified in the latest +C++ mapping for CORBA 3.0 (Document ptc/03-06-03).</li> + +<li> +IDL compiler is now able to generate code that support native C++ exceptions +on the stubs and skeletons. With this strict mapping, the CORBA::Environment +parameter is no longer generated. The default behavior is to generate code +without the extra parameter on plaforms with native exceptions and with +the extra parameter in platforms without native exceptions. Use the -Ge +flag to override the defaults.</li> + +<li> +We are now able to handle shared case labels and default label in unions. +In addition, whenever appropriate, we are also able to generate the "default +()" operation.</li> + +<li> +We are now able to handle recursive types. We are also able to generate +optimized typecodes.</li> + +<li> +Struct members of type strings and arrays of strings now use the managed +type instead of the _var type. This change was necessary to conform to +the IDL->C++ mapping.</li> + +<li> +Fixed a large number of problems with anonymous arrays and sequences inside +structs and unions. The name of anonymous sequence needs to be fixed as +per latest C++ mapping spec.</li> + +<li> +Compile problems with sequence of forward declared interfaces is fixed. +In addition, problems with sequence of CORBA::Objects is fixed. In this +specific case, we were not generating the _downcast and _upcast methods.</li> + +<li> +Some more problems with the front-end have been fixed. In particular, oneway +operations with a "raises" clause or having an "inout", "out", or "return" +mode is flagged as an error.</li> + +<li> +For platforms that support namespaces, we now allow reopening modules.</li> + +<li> +Support for generating compiled marshaling code is added. Use the -Gc option. +However, this needs thorough testing before we can claim success. Unions +are still a problme with compiled marshaling.</li> + +<li> +The problem of "#include"ing the relative path of the header files rather +than the paths of their corresponding IDL files has been fixed. tao_idl +now generates #include path names that are derived from the IDL files that +are #include'd in the main idl file.</li> + +<li> +Added options to IDL compiler to specify file name endings for the IDL-generated +stubs, skeletons and the various header files. Please refer to the <a href="../compiler.html">IDL +compiler options</a> for details.</li> + +<li> +Verified support for the "long long" and "unsigned long long" datatypes. +On platforms that do not support 64 bit longs we provided <i>partial</i> +emulation through ACE_U_LongLong.</li> + +<li> +Perfect Hashed Operation Lookup Strategy has been added to the IDL Compiler. +-P flag to <tt>tao_idl</tt> enables the perfect hased lookup strategy. +This strategy uses <a href="http://www.cs.wustl.edu/~schmidt/gperf.ps.gz">GPERF</a>, +the GNU's Perfect Hash Function Generator written by Dr.Douglas C. Schmidt. +Right now, GPERF works only on Solaris. Any work on porting GPERF to other +platforms will be highly appreciated.</L1></li> + +<li> +The <<= and >>= operators for user-defined types are now generated.</li> + +<li> +Completely redesigned the IDL compiler using the Visitor patterns. Many +incomplete issues have been resolved. These include support for "sequence +of typecodes", passing object references as in, inout, and out parameters. +Code generation for sequences is also properly handled i.e., for a named +sequence such as <tt>typedef sequence<char>CharSeq;</tt>, we now generate +a new class (and hence a type) called "class CharSeq". Arrays are still +being worked out and will be done soon. An important difference in the +generated code is that the skeletons now use a table driven approach very +similar to the stubs.</li> + +<li> +Support for the "native" keyword added.</li> + +<li> +The problem of incorrect code generation for typedefs defined in an imported +file is resolved.</li> + +<li> +Problems when interfaces use single or multiple inheritance solved. The +problem was with the demultiplexing code, the generated operation tables, +and the dispatching mechanism. We are currently testing this with the Event +Channel code.</li> + +<li> +The problems arising due to public virtual inheritance when casting from +an interface class to CORBA::Object_ptr has been solved. We do this casting +inside the stubs/skeletons rather than first converting an interface class +pointer to a void*, storing it in an Any, and casting it to CORBA::Object_ptr +in the encode/decode methods. The casting inside the stubs/skeletons work +because the compiler has knowledge of both types.</li> + +<li> +Include files are handled properly. So are the definitions used inside +the include files that are used in the currently parsed files.</li> + +<li> +Generates C++ stubs and skeletons that use TAO's +<a href="http://www.cs.wustl.edu/~schmidt/HICSS-97.ps.gz"> +interpretive IIOP protocol engine</a>.</li> + +<li> +Support dynamic libraries on NT, i.e., marking classes for DLL export was +added. Two backend options control the name of the export macro, and the +name of an extra include file were the macro is defined; the options are +<tt>-Wp,export_macro=MACRO_NAME-Wp,export_include=INCLUDE_NAME</tt>.</li> + +<li> +The IDL compiler generates now source code for sequences. The user has +now the option to use these generated sequence classes or to use, as up +to now, the template instatiation. If TAO_LACKS_TEMPLATE_SPECIALIZATION +is defined, then template instantiation will be used, else not. The reason +for this was, that some C++ compilers did not support template instantiation +properly and sequences were based on templates. The generated source code +is mainly contained in the generated header file directly in the class +declaration.</li> + +<li> +The IDL Compiler generates templates for servant implementations. The options +are -GI [ h | s | b | e | c ]</li> + +<li> +The IDL compiler generates source code for the management and (de)marshaling +of wide characters and wide strings, enabling the sending and receiving +of Unicode over the wire. However, wide character and wide string literals +are not yet portable to Unix platforms (see entry under Future Work below).</li> + +<li> +Since the CORBA spec requires that all enums be 32 bits, and some compilers +will try to use less space if the enum values are small enough, the IDL +compiler now appends <enum name>_TAO_ENUM_32BIT_ENFORCER = 0xFFFFFFFF +to every enum. This appended enum value is not part of the IDL compiler's +internal representation of the enum, so unions that use the enum as a discriminator +will not have incorrect _default() code generated for them.</li> + +<li> +The IDL compiler generates a C++ ostream operator for IDL exceptions. So +far only the repository ID is output, but this may be enhanced when requirements +and/or desires become clearer.</li> + +<li> +The IDL compiler has support for valuetypes (see release notes +for valuetypes for details). </li> + +<li> +As part of the implementation of interceptors, the TAO IDL compiler now +generates interception points in the client and server, as well as the +prepare_header method in the stubs.</li> + +<li> +Scoping and name resolution rules have changed in CORBA with version 2.3. +The IDL compiler now conforms to these new rules.</li> + +<li> +IDL compiler now supports the CORBA AMI callback model, generating code +for reply handlers and reply stubs if the -GC command line option is used. +The TAO library must be compiled with TAO_HAS_CORBA_MESSAGING = 1. If this +is done, TAO_HAS_AMI_CALLBACK will automatically be defined to 1 as well. +IDL_HAS_VALUETYPE is defined to 1 by default.</li> + +<li> +New command line option -So added to suppress generation of ostream operators +for exceptions.</li> + +<li> +New command line option -Sc added to suppress generation of tie classes +and *S_T.* files. The default is still to generate them.</li> + +<li> +IDL compiler now handles escaped identifiers (CORBA 2.3.1). An identifier +appearing in an IDL file with a leading underscore will appear in generated +code without the underscore. This enables the use of identifiers in generated +code identical to IDL keywords, as specified in CORBA 2.3.1. If the resulting +identifier matches a C++ keyword, "_cxx_" will be prepended in generated +code as before.</li> + +<li> +The -St option to suppress generation of typecodes now also suppresses +the generation of the Any insertion and extraction operators. The extraction +operators require the associated typecode, so the generated code for those +operators would not compile when the -St option was used.</li> + +<li> +Option -Ge 2 added which generates 'throw' instead of ACE_THROW_SPEC, ACE_THROW, +and ACE_RETHROW. Since the expansion of ACE_THROW_RETURN is platform-dependent, +it was left as is. Same for TAO_INTERCEPTOR_THROW, since it sometimes expands +to ACE_THROW_RETURN. Of course ACE_HAS_EXCEPTIONS must be defined for this +option to work.</li> + +<li> +Removed generation of ostream operators for user exceptions, along with +the command line option to suppress this generation. The ostream operator +defined in the base class CORBA::Exception works fine and produces the +same output.</li> + +<li> +The TAO IDL compiler is no longer monolithic. It is composed of a top-level +executable, a front end library and a back end library. This will enable +different back ends to be plugged in for code generation in different languages, +and for IfR administration. Different back ends may require a few changes +to the executable (described below), but the front end library can remain +unchanged. The chain of dependencies is as follows:</li> + +<blockquote>FE : ACE +<br>BE : FE +<br>EXE : FE, BE</blockquote> +Executing the Makefile (or the TAO_IDL Compiler project in the MSVC 'tao_idl' +workspace) will build whatever is required in the proper order, as before. +Back end files, classes and functions required by the executable are as +follows: +<blockquote>be.h - file containing #includes of the major BE file headers. +<br>TAO_Codegen - class, holds output stream references for the various +generated files. +<br>TAO_CODEGEN - ACE_Singleton typedef of the above class. +<br>tao_cg - pointer to instance of TAO_CODEGEN. +<br>be_generator - class, inherits from AST_Generator in FE and generates +AST nodes. +<br>void BE_produce (void) - global function, starts AST traversal for +code generation. +<br>BE_GlobalData - class, holds default/command line arg settings specific +to BE. +<br>be_global - pointer to instance of above class.</blockquote> +All the code in the executable that may need to be modified for different +back ends in contained in the file drv_args.cpp. Code in this file processes +the command line arguments, outputs a usage message, and performs other +miscellaneous BE initialization. +<li> +The IDL compiler can how handle interfaces forward declared in one IDL +file and defined in another. It will handle interfaces that are mutually +dependent across two IDL files, as long as code generated from both IDL +files is included in the same C++ build.</li> + +<li> +Generation of template tie class declarations has been moved from the *S.h +file to the *S_T.h file. The new SunCC 5.2 compiler does not require that +template source code be included in the header file, but it does require +that template declarations and implementations be in the same 'compilation +unit' that is, the same header/source file combination. Generated ifdef +guards prevent the compilation of tie class code if the target platform +does not support namespaces. For such platforms, TAO maps the IDL 'module' +keyword to a class instead of to 'namespace', and C++ does not allow template +class declarations to occur inside another class.</li> + +<li> +IDL compiler can now process multiple IDL files per execution, on all platforms. +A separate process is spawned for each file. IDL files and command line +options may appear in any order on the command line. Any option not starting +with '-' will be treated as a filename.</li> + +<li> +Support for value types has been expanded to include forward declared value +types, sequences of value types, and factory methods. Also, work to support +Anys and type codes for value types is in progress.</li> + +<li> +Support for #pragma prefix has been revamped and improved to be compliant +(the prefix is now cleared when leaving the scope or the included IDL file +where it was defined), and support for #pragma version and #pragma ID has +been added.</li> + +<li> +Support for forward declared structs and unions has been added. In compliance +with CORBA 2.6, such forward declarations must be fully defined in the +same compilation unit, and may be used only in sequence declarations, which +in turn may be used in the declaration of recursive structs and unions. +Use of a forward declared struct or union as an aggregate type member or +as an operation parameter will generat an error message.</li> + +<li> +Support for enum constants has been added.</li> + +<li> +The IDL compiler can now handle concatenation of string literals, for example</li> + +<blockquote>const string foo = "hel" "lo " "the" "re";</blockquote> + +<li> +Support for abstract interfaces has been added.</li> + +<li> +Support for valuetypes has been expanded.</li> + +<ul> +<li> +Forward declared valuetypes not defined in the same compilation unit</li> + +<li> +Type codes for valuetypes</li> + +<li> +Generation of Any insertion/extraction operators</li> + +<li> +Abstract valuetypes</li> + +<li> +Valuetype members of all IDL aggregate types</li> + +<li> +Interface supports list</li> +</ul> + +<li> +Support for the CCM extensions to IDL, which first appeared in CORBA 3.0 +is complete, with the exception of support for the 'import' keyword.</li> + +<li> +Generated code for constants defined in a module is now inlined +(value is assigned in the header file) by default, enabling such +constants to be referenced by name as array bounds, etc. either in +subsequent generated code or in application code. However, this inlining causes +a problem with some compilers when using pre-compiled headers, so the +option -Guc has been added to turn it off. NOTE: Constants +defined at global scope (always inlined) and constants defined inside an +interface or a valuetype (never inlined) are not affected by the option. +</li> + +<li> +Generation of explicit template instantiations is now off by default. The +option -GT turns it on. If turned on, the generated explicit instantiations +are guarded by the same #ifdef preprocessor directives as before. +</li> + +<li> +The mechanism for processing multiple IDL files in a single execution has +changed. Instead of spawning a separate process for each IDL file, the +IDL compiler now processes the files sequentially in a single process. +Note that, for each file, execution of the preprocessor spawns a new +process, as it always has. +</li> + +<li> +Added command line option -Sm to suppress the action of CCM preprocessing +visitor, which is by default launched before the C++ generating visitors, +in order to add CCM equivalent IDL to the AST. There is a new tool in CIAO +which converts IDL files by replacing IDL3 constructs with equivalent IDL2 +constructs, and if the IDL compiler is run on these files, we do not want +the CCM preprocessing to occur. +</li> + +<li> +Changed the behavior of the -o option to create the specified directory +if it does not already exist. Only one level of directory can be created +with this option (any path prefix included must already exist). +If the specified directory already exists, no action is taken. +</li> + +<li> +As part of the subsetting in TAO of Anys and TypeCodes, the generation of +these things in the IDL compiler has been further decoupled from other +code generation. Now the -GA option generates not only a *A.cpp file but +also a *A.h file, and in addition, the -oA option will create these files +in a directory different from that where the other generated files are +created.</li> + +<li> +Generation of explicit template instantiations has been completely +removed, it is no longer available via the command line option -GT.</li> + +<li> +Generation of Any insertion/extracion operators for local interfaces +has been enabled by default. An additional command line option -Sal +has also been added to suppress the generation of Any operators for +local interfaces only.</li> + +<li> +Added new command line option -oS to specify output directory for +all skeleton files, including TIE class files, if generated. Overrides +-o option value, if any.</li> + +<li> +Uninlined all generated TIE class code, since inlining could potentially +cause a problem for RTTI.</li> + +<li> +Added new command line option <tt>-Gse</tt> that causes generation of +an explicit export of each sequence's template base class. This is +occasionally necessary as a workaround for a bug in Visual Studio +(.NET 2002, .NET 2003 and Express 2005) where the template instantiation +used for the base class isn't automatically exported. See +<a href="http://support.microsoft.com/default.aspx?scid=kb;en-us;309801"> +KB 309801</a> for more info</li> + +</ul> + +<h4> +Known Issues:</h4> + +<ul> +<li> +With Microsoft Visual C++, verison 6.0 and earlier, a problem has been +discovered when the IDL compiler is built using the Release version of +the MSVC project. For the Release version, the Optimizations box in the +C/C++ tab has 'Maximize Speed' selected. This setting turned out to cause +a problem when</li> + +<p>const char *foo = ...... +<p>occurs in the source code. The variable foo is sometimes not allocated +or assigned properly, and if foo is part of generated code, it will then +display as garbage or as an empty string (or substring). Two workarounds +have been found. One is to change the declaration of foo to be +<p>static const char *foo = .... +<p>and this has been done in all cases we could find in the IDL compiler +source code. Another way to avoid the problem is to change the Release +project Optimization setting to 'Minimize Size'. Reportedly the problem +has been fixed in MSVC version 7.0. + +</ul> + +<h4> +Future work:</h4> + +<ul> +<li> +Implement the complete <a href="../ReleaseNotes/OBV.html">Object-by-Value</a> +specification. Originally implemented solely for use in holding exceptions +raised in <a href="../../../PDFs/ami1.pdf">asynchronous</a> invocations, +work is now ongoing to extend TAO's OBV implementation until it includes +the complete CORBA value type specification. Some of the items yet to be +implemented include:</li> + +<ul> +<li> +Boxed value types</li> + +<li> +Custom marshaling</li> + +<li> +Truncation</li> + +<li> +Marshaling of complex state (graphs with cycles)</li> + +</ul> +For a complete description of value type semantics, see chapter 5 in the +CORBA specification. +<li> +The generated sequence classes should not be generated per sequence, but +per type and parent scope. Which means, that the overhead of having the +source code generated serveral times should be reduced. To do this, an +extra pass over the internal representation of the IDL file has to be done.</li> + +<li> +Updated and portable support for wide characters and wide strings. The +original implementation supports Unicode only. At the time, the CORBA specification +required wide characters, whether standalone or in a wide string, to be +marshaled as 16-bit quantities. This is the same size as the native wchar_t +type on Win32 platforms. However, the wchar_t type on UNIX and related +platforms is 32 bits in size. In addition, the General Inter-ORB Protocol +(GIOP) version 1.2 has changed the CDR transfer syntax for wide characters. +In the new version, a wide character is marshaled as a single byte indicating +the size of the wide character representation, followed by the specified +nunber of bytes. The IDL compiler's handling of wide characters needs to +be updated, both to handle the GIOP 1.2 CDR transfer syntax, and to be +portable under previous versions of GIOP. For more information about GIOP +CDR transfer syntax, see section 15.3 in the CORBA specification. and for +information about the marshaling of IDL character types specifically, see +section 15.3.1.6.</li> + +</ul> + +<p><br><!--#include virtual="orbcore.html" --> +<hr> +<h3> +<a NAME="pp"></a>Pluggable Protocols</h3> +Point of contact: <a href="mailto:ossama@dre.vanderbilt.edu">Ossama Othman</a> +<p>The goal of the pluggable protocol effort is to (1) identify logical +communication layers in the ORB, (2) abstract out common features, (3) +define general interfaces, and (4) provide necessary mechanisms for implementing +different concrete ORB and transport protocols. TAO's pluggable protocol +framework will allow disparate communication mechanisms to be supported +transparently, each with its own set of requirements and strategies. +<p>For example, if the ORB is communicating over a system bus, such as +PCI or VME, and not all the features of GIOP/IIOP are necessary and a simpler, +optimized ORB and transport protocol can be defined and implemented. Similarly, +it should be straightforward to add support for new transport protocols +that use native ATM or shared memory as the underlying communication mechanism. +In all cases the ORB's interface to the application will remain compliant +with the OMG CORBA standard. +<p>There will be several stages of the development process: (1) basic pluggable +transport protocols framework, (2) support for multiple profiles, (4) add +example transport protocols, such as ATM and VME, and refine/optimize the +transport protocols framework, and (4) add support for pluggable ORB protocols, +e.g., replacements for GIOP. Each of these steps is outlined below: +<ul> +<li> +<b>Basic pluggable transport protocols framework</b>: We have added several +Bridge classes that decouple the transport-specific details from the rest +of TAO's ORB Core. This allows us to isolate the details of how messages +are communicated at the transport layer in a few classes. This design resulted +in the restructuring of the ORB Core and how requests are handled. For +instance, there is now the concept of communication layers: Objects (e.g., +references, method invocations, etc.), ORB Messaging, Transport, and Network. +The Object layer is just the usual stubs and skeletons.</li> + +<p>The common interfaces have been defined in the new abstract classes +that form the core of TAO's pluggable protocol framework, e.g., +<tt>TAO_Connector</tt>, +<tt>TAO_Acceptor</tt>, +<tt>TAO_Profile</tt> +and <tt>TAO_Transport</tt>. Two new mechanisms for keeping track of supported +transport protocols are the +<tt>TAO_Connector_Registry</tt> and +<tt>TAO_Acceptor_Registry</tt>, +which are essentially Abstract Factories that produce the right types of +connector, acceptors, and transports. +<li> +<b>Multiple Profile</b> - Support for more than one profile per object. +This is important since there may be several different ways to access an +object. Each profile for an object may encode information pertaining to +QoS, network and transport protocols, addresses or routes.</li> + +<li> +<b>Example Transport protocols</b> - Aside from IIOP, the following transport +protocols are distributed with TAO:</li> + +<ol> +<li> +UIOP: GIOP over local IPC (UNIX domain sockets)</li> + +<li> +SHMIOP: GIOP over shared memory</li> + +<li> +SSLIOP: GIOP over SSL (Secure Socket Layer)</li> + +<li> +SCIOP: GIOP over SCTP</LI> + +<li> +COIOP: GIOP collocated only. Will only work in a collocated environment, no remote calls are possible</li> + +<li> +DIOP: GIOP over UDP/IP unicast</li> + +<li> +MIOP: GIOP over UDP/IP multicast</li> +</ol> +Other interesting transport protocols could be for ATM, Buses (VME or PCI), +TP4, and GSMP. TAO users have also created their own pluggable transport +protocols, such as a ScramNet pluggable protocol. +<li> +<b>Pluggable ORB protocols</b> - This step will add support for ORB protocols +besides GIOP. In particular, we will explore lightweight protocols using +shared memory and system buses like PCI or VME.</li> +</ul> +Current Status: +<ul> +<li> +The basic framework to support pluggable transport protocols has been completed. +The standard TAO regression tests <tt>Latency</tt>, +<tt>MT_Cubit</tt>, +<tt>Multiple_Inheritance</tt>, +<tt>CDR</tt> and +<tt>EC_Throughput</tt> can be used to verify performance +using the new framework.</li> + +<li> +Multiple endpoint support in the ORB has been added. A list of TAO_Acceptors +is kept in the Acceptor Registry. When the ORB needs to create an IOR it +iterates over all the acceptors to do so. Using either multiple <tt><a href="../ORBEndpoint.html">-ORBEndpoint</a></tt> +options or several endpoints separated by semi-colons ';', the user can +specify what addresses the ORB should use. Each endpoint is specified in +URL format (ex: <tt>iiop://foo.bar.com:0</tt>), this format can be extended +to support different protocols.</li> + +<li> +If the user does not specify a list of endpoints then the ORB creates a +default endpoint for each protocol configured, unless the pluggable protocol +explicitly prevents that in an effort to prevent resource leaks.</li> + +<li> +Added support for multiple Connectors in the ORB, the ORB finds the correct +connector based on the tag for the profile.</li> + +<li> +Added support for multiple profiles in the IORs, when the ORB demarshals +an IOR it queries the Connector Registry to create the right kind of profile +for the known protocols. If one of the protocols is unknown we create a +special profile class that can only be used for marshaling and demarshaling, +not communication.</li> + +<li> +Enabled the UIOP protocol. This protocol uses local IPC (aka UNIX domain +sockets) as the transport mechanism. The protocol is loaded by default.</li> + +<li> +Enabled the SHMIOP protocol. This protocol uses shared memory as the transport +mechanism. The protocol is loaded by default.</li> + +<li> +An IIOP over SSL pluggable transport called "SSLIOP" has been implemented. +It provides secure communication between hosts that support IIOP over SSL, +and is meant to be a drop-in replacement for the IIOP pluggable transport. +TAO's SSLIOP pluggable transport implementation supports both the standard +IIOP transport protocol and the secure IIOP over SSL transport protocol.</li> + +<p>No changes were made to the core TAO sources to provide to this SSL +support, nor does TAO contain any security related hooks. TAO's SSLIOP +implementation is completely self-contained. This ensures that the core +TAO sources remains free of export restrictions. +<li> +Protocols can be dynamically loaded into the ORB: The default resource +factory reads the protocol "names" from its list of arguments. These protocol +names are used to load an abstract factory via the service configurator. +This factory can create acceptors or connectors on demand. By default only +IIOP is loaded.</li> + +<li> +The service configurator is now used to load protocol factories.</li> + +<li> +Support for the <tt>-ORBHost</tt> and <tt>-ORBPort</tt> has been removed. +The new <tt><a href="../ORBEndpoint.html">-ORBEndpoint</a></tt> option +supersedes them, and provides the same functionality in a protocol-neutral +way. If the deprecated options are used, the ORB exits with a <tt>CORBA::BAD_PARAM</tt> +exception, indicating an unknown <tt>-ORB</tt> option.</li> + +<li> +The <tt>-ORBPreconnect</tt> ORB option has been deprecated in favor of +the standard <tt>validate_connection</tt> run-time feature. Support for +this option will be removed from future releases.</li> + +<li> +The URL style object reference format has been updated to conform with +the format that <tt>corbaloc</tt> uses. The BNF specification for <tt>corbaloc</tt> +is:</li> + +<blockquote><tt><corbaloc> = "corbaloc:/"[<obj_addr_list>]["/"<key_string>]</tt> +<br><tt><obj_addr_list>= [<obj_addr> ","]* <obj_addr></tt> +<br><tt><obj_addr>= <prot_addr> | <future_prot_addr></tt> +<br><tt><prot_addr>= <rir_prot_addr> | <iiop_prot_addr></tt> +<br><tt><rir_prot_addr>= <rir_prot_token>":"</tt> +<br><tt><iiop_prot_addr>= <iiop_id><iiop_addr></tt> +<br><tt><iiop_id>= ":" | <iiop_prot_token>":"</tt> +<br><tt><iiop_prot_token> = "iiop"</tt> +<br><tt><iiop_add> = [<version> <host> [":" <port>]]</tt> +<br><tt><host> = DNS-style Host Name | ip_address</tt> +<br><tt><version> = <major> "." <minor> "@" | empty_string</tt> +<br><tt><port> = number</tt> +<br><tt><major> = number</tt> +<br><tt><minor> = number</tt> +<br><tt><future_prot_addr> = <future_prot_id><future_prot_addr> +<future_prot_id> = <future_prot_token>":" <future_prot_token> += possible examples: "atm" | "dce" <future_prot_addr> = protocol specific +address <key_string> = <string> | empty_string</tt></blockquote> +The <tt>uiop</tt> URL style object references syntax is: +<blockquote><tt><uioploc> = "uioploc://"[<addr_list>]["|"<key_string>]</tt> +<br><tt><addr_list>= [<address> ","]* <address></tt> +<br><tt><address> = [<version> <rendezvous point>]</tt> +<br><tt><rendezvous point> = Valid Filesystem Path</tt> +<br><tt><version> = <major> "." <minor> "@" | empty_string</tt> +<br><tt><major> = number</tt> +<br><tt><minor> = number</tt> +<br><tt><key_string> = <string> | empty_string</tt></blockquote> +Note that the key string delimiter for <b><tt>uiop</tt></b> is a vertical +bar `<b><tt>|</tt></b>' (the command line "pipe" symbol) not a forward +slash `<tt>/</tt>'. A delimiter other than a forward slash is needed to +prevent ambiguities of where the rendezvous point ends and where the key +string begins since both may contain forward slashes in them. The new <tt>corbaloc:uiop</tt> +URL format may also be used. +<p>It should be noted that these formats have been superseded by the new +<tt>corbaloc</tt> +<a href="../INS.html">Interoperable Naming Service</a> support. +<li> +The <i>rendezvous point</i> for <tt>uiop</tt> is any valid path and filename +that the ORB has permission to read and write to. However, UIOP rendezvous +points have the same restrictions that local IPC has. The following are +some guidelines that will help ensure successful use TAO's UIOP pluggable +transport protocol:</li> + +<blockquote> +<li> +To guarantee portability, local IPC rendezvous points (including the path +and filename) should not be longer than 99 characters long. Some platforms +may support longer rendezvous points, usually 108 characters including +the null terminator, but Posix.1g only requires that local IPC rendezvous +point arrays contain a maximum of <b>at least</b> 100 characters, including +the null terminator.</li> + +<P>If an endpoint is longer than what the platform supports then +it will be truncated so that it fits, and a warning will be issued. +<li> +Avoid using <i>relative</i> paths in your UIOP endpoints. If possible, +use <b><i>absolute</i></b> paths instead. Imagine that the server is given +an endpoint to create using <tt>-ORBEndpoint uiop://foobar</tt>. A local +IPC rendezvous point called <tt>foobar</tt> will be created in the current +working directory. If the client is not started in the directory where +the <tt>foobar</tt> rendezvous point exists then the client will not be +able to communicate with the server since its point of communication, the +rendezvous point, was not found. On the other hand, if an absolute path +was used, the client would know exactly where to find the rendezvous point.</li> + +<p>It is up to the user to make sure that a given UIOP endpoint is accessible +by both the server and the client. +<li> +It is important to be consistent in the use of absolute paths and relative +paths for rendezvous points. The two types of paths should not be used +for the same endpoint. For example, if <tt>uiop:///tmp/foo</tt> is specified +as the server endpoint and <tt>uiop://foo</tt> as a preconnect for a client +in <tt>/tmp</tt>, then the preconnection may be established but it is likely +it won't be used since the endpoint and preconnect are interpreted as different +strings, i.e. <tt>/tmp/foo</tt> and <tt>foo</tt> are not the same, lexicographically. +On the other hand, if both the endpoint and the preconnect are the same +string then a preconnection will be established and used successfully.</li> + +</blockquote> +The <tt><a href="../ORBEndpoint.html">-ORBEndpoint</a></tt> option uses +a syntax similar to that of the URL style object reference shown above. +The only difference is that the object key delimiter and the object key +string are not specified. +<li> +Added documentation that describes how to implement pluggable transport +protocols for TAO. The document is available <a href="../pluggable_protocols/index.html">here</a>.</li> + +<li> +TAO's IIOP pluggable protocol now supports automatic creation of profiles +for endpoints created on a host with multiple network interfaces. It should +no longer be necessary to manually specify an endpoint for each network +interface.</li> + +<p>This means that server IORs will contain profiles for all of the default +endpoints created on each network interface the server is listening on, +if no explicit endpoints were specified.</ul> +Known Issues: +<ul> +<li> +</li> +</ul> +Critical Work: +<ul> +<li> +None.</li> + +</ul> +Future Work: +<ul> +<li> +Complete support for multiple ORB messaging protocols.</li> + +<li> +Long term work will include adding support for pluggable ORB protocols, +as well as transport protocols. This way we can develop optimal messaging +and transport protocols for a given platform.</li> +</ul> + +<hr> +<h3> +<a NAME="poa"></a>Portable Object Adapter (POA)</h3> +Point of contact: <a href="mailto:irfan@cs.wustl.edu">Irfan Pyarali</a> +<p>The POA associates servants with the ORB and demultiplexes incoming +requests to servants. +<p>Current Status: +<ul> +<li> +TAO supports the POA spec. This section will carry updates as available.</li> +</ul> +Known issues: +<br>Future work: +<br>Recently completed work: +<ul> +<li> +ORB::shutdown now properly deactives all the POA Managers.</li> + +<li> +POA Managers in TAO were previously ignored in the request processing path +on the server. This is now fixed such that their state is checked before +dispatching the client request to the servant. Only if the state is <tt>ACTIVE</tt>, +is the request dispatched to the servant. Otherwise, the request is rejected. +Since POA Managers start off in <tt>HOLDING</tt> state, make sure to +<tt>activate()</tt> +them before falling into the event loop.</li> + +<li> +TAO's POA now properly supports both the threading policies: SINGLE_THREAD_MODEL +and ORB_CTRL_MODEL.</li> + +<li> +The synchronization in the POA is now very optimal. For example, the locks +are not held across the invocation on the servant. The locks are also not +held across the invocation on the AdapterActivator and ServantManagers. +This allows us to use regular locks instead of recursive locks inside the +POA. This also allows multiple threads to dispatch requests on the same +POA simultaneous.</li> + +<li> +Before 1.4.6 TAO supports reference counting between POA and servants, including +the new RefCountServantBase and ServantBase_var classes. RefCountServantBase +is a reference counted base class that was added to the CORBA specification +to avoid race conditions for servant deletion in threaded servers. <a href="ftp://ftp.omg.org/pub/docs/orbos/98-07-12.pdf">ftp://ftp.omg.org/pub/docs/orbos/98-07-12.pdf</a> +contains the relevant text. Check <a href="../poa_migration.html#Reference counting Servants">here</a> +on some hints to avoid trouble. From TAO 1.4.6 reference counting is always +enabled for servants and RefCountServantBase is a noop struct you don't +need anymore.</li> + +<li> +The POA now supports active demultiplexing of servants in the SYSTEM_ID +and the USER_ID policy. This should make the POA faster and more predictable +since there is no hashing involved and the index of the slot where the +servant is registered is in the Object Key.</li> + +<li> +Previously, the complete POA name was used as the POA identity. This scheme +was inefficient in many ways including: (a) the complete POA name can be +significantly large in size, and therefore, ineffient to pass with every +method call from the client to the server; (b) it is varible in size, and +therefore, does not lend itself to smart and effective parsing; (c) the +searching based on the complete POA name is very ineffient.</li> + +<p>The new solution is to use an active demux table, and flatten the POA +hierarchy. This will help in the searching since active demuxing is fast +and predictable. This will also help in the parsing since the demux key +will be fixed size. +<p>Note that for persistent ids, we have to pass the complete POA name +in addition to the demux key in order to handle POA creation on demand. +<li> +There were some POA objects in a typical server that are not freed up properly, +resulting in a memory leak. This has now been fixed.</li> + +<li> +Timestamps in persistent IORs were not required and have been removed.</li> + +<li> +POA exceptions are not not system exceptions and have been removed from +the list of system exceptions.</li> + +<li> +Vastly improved the ability of the POA to deal with user exceptions, memory +allocation failures, and constructor failures.</li> + +<li> +We now support a minimum POA for the minimum CORBA specification. Recently, +this feature was enchanced such that the minimum CORBA footprint was further +reduced. In addition, minimum POA can be enabled/disable irrespective of +the minimum CORBA setting.</li> + +<li> +We have decided not to support active demuxing for method name lookup. +The benefit of this optimization was questionable since the current perfect +hashing scheme provide very good and predictable behavior.</li> + +<p>Also, note that this optimization will require many changes. We would +have to use the help of the IDL compiler to modify the object key that +is passed for every method call differently. Note that this scheme doesn't +work in the case of multiple inheritance or when the client stubs are not +TAO. +<li> +Improved the parsing of object keys belonging to the RootPOA. Since this +is the default POA and is commonly used, we have given it a reserved byte +in the object key in order to quickly identify it. With the reserved bit, +the active demux key for the RootPOA is not used, and no map lookups are +required.</li> + +<li> +POA name separator was changed from '/' to '\0'. Since POA names are strings, +this makes a better choice since there is no chance of a conflict with +the string specified by the user.</li> + +<li> +We have support for reactivating servants with system generated ids.</li> + +<li> +The TAO specific synchronization POA policy has been removed.</li> + +<li> +New examples have been added to show how servants can be dynamically loaded +from DLLs on demand.</li> + +<li> +Support for collocation should be much better now because the POA can tell +if we created the object reference.</li> + +<li> +After Nanbor's recent changes for collocation, we support the full semantics +of remote objects on a collocated object. The spec mandates that collocated +object should behave <i>exactly</i> like remote objects, which includes +going through the POA, running the Servant Managers, running the interceptors, +and expecting the reference counting behavior provided by the POA. Note +that the old scheme of direct call through to the servant is also still +available.</li> + +</ul> + +<hr> +<br><!--#include virtual="OBV.html" --> +<h3> +<a NAME="nservices"></a>CORBA Naming Service and Interoperable Naming Service</h3> +Points of contact: <a href="mailto:marina@cs.wustl.edu">Marina Spivak</a> +and <a href="mailto:vishal@cs.wustl.edu">Vishal Kachroo</a> +<p>OMG defined CORBA Naming Service (spec <a href="ftp://ftp.omg.org/pub/docs/formal/97-07-12.pdf">here</a>) +to provide a basic service location mechanism for CORBA systems. CosNaming +manages a hierarchy of name-to-object-reference mappings. Anything, but +typically the server process hosting an object, may bind an object reference +with a name in the Naming Service by providing the name and object reference. +Interested parties (typically clients) can then use the Naming Service +to resolve a name to an object reference. +<p>More recently, CORBA Naming Service was subsumed/extended by the CORBA +Interoperable Naming Service, a.k.a. INS (spec <a href="ftp://ftp.omg.org/pub/docs/orbos/98-10-11.pdf">here</a>). +INS inherits all the functionality from the original Naming Service specification +in addition to addressing some its shortcomings. In particular, INS defines +a standard way for clients and servers to locate the Naming Service itself. +It also allows the ORB to be administratively configured for bootstrapping +to services not set up with the orb at install time. <a href="../INS.html">This +page</a> provides a brief description of additional features INS provides, +and how they are implemented in TAO. +<p>Current status (as of 18 May 2001): +<ul> +<li> +The Naming Service now <i>does not</i> do listen for multicast discovery +as the default. Use <tt>-m1</tt> option to turn it on.</li> +</ul> +Recent work: +<ul> +<li> +Changed default for multicast discovery to</li> +</ul> +Features: +<ul> +<li> +Added support for Persistence (using memory-mapped files). Persistence +feature is optional, and is controlled by the command line argument.</li> + +<li> +Updated the implementation of the Naming Service to use new ACE exception +macros.</li> + +<li> +Added support for the InterOperable Naming Service, which enables the ORB +to support IORs in user-friendly <tt>corbaloc</tt> format. These features +allow the ORB to be configured to return arbitrary object references from +<tt>CORBA::ORB::resolve_initial_references</tt> +for non-locality-constrained objects. Two options -ORBInitRef and -ORBDefaultInitRef +have been added to the orb for bootstrapping to arbitrary services.</li> + +<li> +Added support for the Naming service to act like an agent: to understand +IIOP request messages from clients and respond with reply messages with +a <tt>LOCATION_FORWARD/OBJECT_NOT_EXIST</tt> status. The Naming Service +can be configured through ORB options to register arbitrary services given +the URL-format IOR for the service. The <tt>resolve_initial_references +()</tt> resolves a service in the following order :</li> + +<ol> +<li> +<tt>-ORBInitRef</tt></li> + +<li> +<tt>-ORBDefaultInitRef</tt></li> + +<li> +Multicast discovery</li> +</ol> + +<li> +Added a test for the InterOperable Naming Service that works in conjunction +with the current TAO examples.</li> + +<li> +Implementation of the CORBA INS spec is completed. The following features +are now supported:</li> + +<ul> +<li> +corbaname format</li> + +<li> +NamingServiceExt interface</li> +</ul> +</ul> +Future work: +<ul> +<li> +Support for a load balancing feature similar to the one present in ORBIX. +It will be possible to bind a group of objects under a single name, and +when a client attempts to resolve the name in question, a preset policy +(e.g., random, round robin, etc.) will determine which one of the object +references from the group will be returned.</li> + +<li> +Support for the Naming Service to handle the IIOP LocateRequest messages +and respond with LocateReply messages with a +<tt>LOCATION_FORWARD/OBJECT_NOT_EXIST</tt> +status.</li> +</ul> + +<hr> +<h3> +<a NAME="tservices"></a>CORBA Trading Service</h3> +Point of contact: <a href="mailto:sbw1@cs.wustl.edu">Seth Widoff</a> +<p>The <a href="http://www.omg.org/technology/documents/formal/trading_object_service.htm">Trading +Service</a> is an implementation of the COS Trading Service speficiation +that meets the Linked Trader conformance criteria --- it implements the +<tt>Lookup</tt>, +<tt>Register</tt>, <tt>Admin</tt>, and +<tt>Link</tt> interfaces, but not +the <tt>Proxy</tt> interface. Notably, the TAO trader supports the following +features: +<ul> +<li> +Multithreaded operation;</li> + +<li> +Trader federations and distributed queries;</li> + +<li> +Dynamic properties;</li> + +<li> +Modifiable properties;</li> + +<li> +All policies described in the specification;</li> + +<li> +Preference sorting;</li> + +<li> +Service type inheritance hierarchies and subtype searching.</li> +</ul> +<a href="trader.html">Trading Service documentation</a> is also available. +<p>Future Work: +<ul> +<li> +The Proxy Interface.</li> + +<li> +Persistent storage of service types and offers.</li> +</ul> + +<hr> +<h3> +<a NAME="pservices"></a>CORBA Property Service</h3> +Point of contact: <a href="mailto:alex@cs.wustl.edu">Alexander Babu Arulanthu</a> +<p>Current status (as of Mar 9th, 1999): All the interfaces of this service +have been implemented. Please go through the test examples at $TAO/orbsvcs/tests/CosPropertyService. +Property Service is has been used by the TAO's <a href="#av">Audio Video +Streaming Service</a>developed for TAO. For general documentation of the +Property Service, please read <a href="http://www.omg.org/technology/documents/formal/property_service.htm">The +Property Service Specification.</a> +<p>Recent Work: +<ul> +<li> +Changed the PropertyException from Exception to struct, according to the +OMG's changes.</li> + +<li> +Changed the implementation to allocate storage for the Sequence out parameters, +eventhough their length is 0. This is according to the CORBA specification.</li> +</ul> + +<hr> +<h3> +<a NAME="cservices"></a>CORBA Concurrency Service</h3> +Point of contact: <a href="mailto:tworm@cs.wustl.edu">Torben Worm</a> +<p>Current status (as of May 3rd): The <a href="http://www.omg.org/technology/documents/formal/concurrency_service.htm">Concurrency +Service</a> provides a mechanism that allows clients to acquire and release +various types of locks in a distributed system. +<ul> +<li> +A simple version of the Concurrency Service has been implemented, i.e. +a version without transactions. It is currently being tested.</li> +</ul> +Future Work: +<ul> +<li> +Implementation of the Concurrency Service with transactions</li> +</ul> + +<hr WIDTH="100%"> +<h3> +<a NAME="av"></a>CORBA Audio/Video Streaming Service</h3> +Point of contact: <a href="mailto:yamuna@oomworks.com">Yamuna Krishnamurthy</a> +<p>This is an implementation of the OMG spec addressing the <a href="http://www.omg.org/technology/documents/formal/audio.htm">Control +and Management of Audio/Video Streams</a>. For more documentation on TAO's +A/V Service please have a look <a href="http://www.cs.wustl.edu/~schmidt/av.html">here</a>. +<h4> +Current Status:</h4> +(as of July 30 2002) +<ul> +<li> +The audio/video streaming service has been implemented in the full profile. +The current implementation support all the flow related components like +flowEndpoint,FDev,FlowConnection,..,etc.</li> + +<li> +Point-to_Point and Point-to-MultiPoint streams have been implemented.</li> + +<li> +A Pluggable protocols framework has been implemented to flexibly add new +flow protocols like SFP, RTP and new transports like ATM. The current implementation +has protoocol implementations for SFP, RTP over UDP and Multicast UDP. +Please look at <tt><a href="../../orbsvcs/orbsvcs/AV/AV_Pluggable_Framework.html">orbsvcs/orbsvcs/AV/AV_Pluggable_Framework.html</a></tt>for +more documentation about the implementation.</li> + +<li> +QoS_UDP pluggable protocol added which implements RSVP QoS support for +UDP streams.</li> + +<li> +Diffserv support added to the UDP pluggable protocol by <a href="mailto:crodrigu@bbn.com">Craig +Rodrigues</a>.</li> + +<li> +RTP/UDP support enhancements and RTCP control protocol support added by +<a href="mailto:rruff@scires.com">Rob +Ruff</a>.</li> +</ul> +Work in progress: +<ul> +<li> +Integration with <a href="http://www.dist-systems.bbn.com/projects/AIRES/UAV/">BBN +UAV software.</a></li> +</ul> + +<hr> +<p><a NAME="ts"></a><b>CORBA Time Service</b> +<p>Point of contact: <a href="mailto:vishal@cs.wustl.edu">Vishal Kachroo</a> +<p>The <a href="ftp://ftp.omg.org/pub/docs/formal/97-02-22.pdf">Time Service</a> +allows clients to connect to Time Service Clerks and obtain globally synchronized +time. This time is calculated from the time obtained from one or more Time +Servers running on multiple machines in the network. The service uses the +TAO Implementation Repository to activate the time servers on demand. +<p>Current status (as of 10th Jan 1999): +<ul> +<li> +Implementation of a Distributed CORBA Time Service is complete.</li> +</ul> +Future work: +<ul> +<li> +Currently the average of the time obtained from the various servers is +considered the global notion of time. A better distributed time synchronization +algorithm can be used in the future.</li> + +<li> +Implementation of the Timer Event Service.</li> +</ul> + +<hr WIDTH="100%"> +<h3> +<a NAME="ec"></a>CORBA Event Service</h3> + +<h4> +Last updated: Fri Mar 5 20:38:26 CST 1999</h4> +Point of contact: <a href="mailto:pradeep@cs.wustl.edu">Pradeep Gore</a> +<p>The COS compliant Event Service implements the Event Service Specification: +<a href="ftp://ftp.omg.org/pub/docs/formal/97-12-11.pdf">(.pdf)</a>, +<a href="ftp://ftp.omg.org/pub/docs/formal/97-12-11.ps">(.ps)</a> +<br>The different command line and service configurator options used +for configuring the CORBA event services are located <A +href="../cec_options.html"> here</a>. This implementation is based on +the Real Time Event service. +<h3> +Features in this release:</h3> + +<ul>There is a new implementation of the COS Event Service available. This +new implementation supports both the Push and Pull styles for event communication, +and it does not require the Real-time Event Service to work. +<p>A new testsuite for the COS Event Service has been started, they are +available at: <tt>$TAO_ROOT/orbsvcs/tests/CosEvent/</tt> +<p>A new example for the COS Event Service is provided in <tt>$TAO_ROOT/orbsvcs/examples/CosEC/Simple/</tt> +<p>A new binary to run the native COS Event Service was added, it is compiled +in: <tt>$TAO_ROOT/orbsvcs/CosEvent_Service/CosEvent_Service_Native</tt> +<li> +A simple test (<tt>$TAO_ROOT/orbsvcs/tests/CosEC_Basic</tt>) demonstrates +how to create and use the event channel.</li> + +<li> +Event Service (<tt>$TAO_ROOT/orbsvcs/CosEvent_Service</tt>)The Event Service +creates a COS compliant event channel and registers it with the naming +service with the default name "CosEventChannel".</li> + +<br>Please read the associated README for more details. +<li> +CosEC_Multiple: <tt>($TAO_ROOT/orbsvcs/tests/CosEC_Multiple)</tt>: This +test demonstrates how multiple CosEC's connect to one RtEC and how multiple +consumers and producers exchange events in this configuration.</li> +</ul> + +<h3> +Known bugs:</h3> + +<ul> +<li> +CosEC_Multiple: <tt>($TAO_ROOT/orbsvcs/tests/CosEC_Multiple)</tt>: Once +the tests are done, the control doesn't return to the shell, you have to +say CTRL-C to get back to the prompt.</li> +</ul> + +<hr WIDTH="100%"> +<h3> +<a NAME="log"></a>CORBA Telecom Log Service</h3> + +<h4> +Last updated: Sun May 28 15:42:44 PDT 2006</h4> +Point of contact: <a href="mailto:jtc@acorntoolworks.com">J.T. Conklin</a> +<p>The CORBA <a href="ftp://ftp.omg.org/docs/formal/03-07-01.pdf">Telecom +Log Service</a> was updated in TAO 1.5. +<h3> +Features supported in the current version:</h3> + +<ul> +<li> +The Log Service implementation under <tt>$TAO_ROOT/orbsvcs/orbsvcs/Log</tt> +implements the <tt>DsLogAdmin </tt>module.</li> + +<li> +Support for the <tt>DsEventLogAdmin </tt>module, which uses the <a href="#ec">COS +Event Service</a> has been added.</li> + +<li> +Support for the <tt>DsNotifyLogAdmin </tt>module, which uses the <a href="#notify">Notification +Service</a> has been added.</li> + +<li> +Support for the <tt>RTEventLog </tt>module, which uses the <a href="ec.html">RTEvent +Service</a> has been added.</li> + +<li> +Support for the <tt>DsLogNotification </tt>module for log- generated events +has been added.</li> + +<li> +The Logging_Service (<tt>$TAO_ROOT/orbsvcs/Logging_Service</tt>) contains +4 separate services</li> + +<ul> +<li> +Basic_Logging_Service <tt>$TAO_ROOT/orbsvcs/Logging_Service/Basic_Logging_Service</tt></li> + +<li> +Event_Logging_Service <tt>$TAO_ROOT/orbsvcs/Logging_Service/Event_Logging_Service</tt></li> + +<li> +Notify_Logging_Service <tt>$TAO_ROOT/orbsvcs/Logging_Service/Notify_Logging_Service</tt></li> + +<li> +RTEvent_Logging_Service <tt>$TAO_ROOT/orbsvcs/Logging_Service/RTEvent_Logging_Service</tt></li></ul> + +<li> +Each service registers with the Naming Service as +"BasicLogFactory", "EventLogFactory", "NotifyLogFactory" and +"RTEventLogFactory" respectively.</li> + +<li> +Each service registers with the Interoperable Naming Service as +"BasicLogService", "EventLogService", "NotifyLogService" and +"RTEventLogService" respectively.</li> + +<li> +The Log Service uses a dynamically loaded "plug-in" Strategy for storing and +querying log records. The default Strategy stores Log records in memory and +supports the Extended Trader Constraint Language (ETCL) Query Language.</li> + +<li> +There are now examples that demonstrate simple usage of the Log Services. +These are found in <tt>$TAO_ROOT/orbsvcs/examples/Log</tt> and the relevant +sub-directories.</li> + +<br>A thorough test <tt>$TAO_ROOT/orbsvcs/tests/Log/Basic_Log_Test</tt> +tests most of the features of the basic logging service.</li> +</ul> + +<h3> +Future work and enhancements:</h3> + +<ul> +<li> +Provide Strategies that support persistent storage of log records.</li> + +<li> +Change default Strategy to use Red-Black trees to optimize lookup on +frequently used query keys - namely record id's and time.</li> +</ul> + +<hr WIDTH="100%"> +<h3> +<a NAME="notify"></a>Notification Service</h3> + +<h4> +Last updated:Thu Nov 21 18:41:11 2002</h4> +Point of contact: <a href="mailto:pradeep@oomworks.com">Pradeep Gore</a> +<p>TAO's CORBA <a href="ftp://ftp.omg.org/pub/docs/telecom/99-07-01.pdf">Notification +Service </a>implementation consists of the following (see the associated +README's for more information): +<ul> +<li> +The implementation of the interfaces in the Notification Service spec is +in <tt><a href="../../orbsvcs/orbsvcs/Notify">$TAO_ROOT/orbsvcs/orbsvcs/Notify +</a></tt>.</li> + +<li> +The service driver is implemented in <tt><a href="../../orbsvcs/Notify_Service">$TAO_ROOT/orbsvcs/Notify_Service</a></tt>.</li> + +<br>The various options of the Notify_Service are described <a href="notify.html#notify_Service">here +</a>. +<li> +The example in <tt><a href="../../orbsvcs/examples/Notify/Filter">$TAO_ROOT/orbsvcs/examples/Notify/Filter</a></tt> +shows a basic example of using filters.</li> + +<li> +The example in <tt><a href="../../orbsvcs/examples/Notify/Subscribe">$TAO_ROOT/orbsvcs/examples/Notify/Subscribe</a></tt>shows +a basic example of how to use subscriptions.</li> + +<li> +Feature unit tests are under <tt><a href="../../orbsvcs/tests/Notify">$TAO_ROOT/orbsvcs/tests/Notify</a></tt></li> + +<li> +The <a href="notify.html#feature_matrix">Notification feature matrix </a>lists +the features implemented.</li> +</ul> +<i>Note that this implementation does not support Pull interfaces and Typed +Event style communication.</i> +<p> +<hr WIDTH="100%"> +<h3> +<a NAME="rt_notify"></a>Real-Time Notification Service</h3> + +<h4> +Last updated:Thu Jul 24 11:57:53 2003</h4> +Point of contact: <a href="mailto:pradeep@oomworks.com">Pradeep Gore</a> +<p>This is an extension to TAO's CORBA Notification Service with Real-Time +CORBA support. +<p><b>IDL Extensions:</b> +<ul> +<li> +<b>Interface Extensions:</b> The <a href="../../orbsvcs/orbsvcs/NotifyExt.idl">$TAO_ROOT/orbsvcs/orbsvcs/NotifyExt.idl</a> +extends the ConsumerAdmin and SupplierAdmin interfaces. The <tt>obtain_notification_push_supplier_with_qos</tt> +and <tt>obtain_notification_push_consumer_with_qos</tt> methods can be +used to specify QoS parameters.</li> + +<li> +<b>QoS Definitions:</b></li> + +<ul> +<li> +ThreadPoolParams: This specifies the parameters for creating a threadpool +in an RT POA.</li> + +<li> +ThreadPoolLanesParams: This specifies the parameters for creating a threadpool +with lanes in an RT POA.</li> +</ul> +</ul> +<b>Tests and Examples:</b> +<ul> +<li> +The <tt><a href="../../orbsvcs/examples/Notify/Lanes">$TAO_ROOT/orbsvcs/examples/Notify/Lanes +</a></tt>example +shows how to use RTCORBA Lanes in RT Notification.</li> + +<li> +The <tt><a href="../../orbsvcs/examples/Notify/ThreadPool">$TAO_ROOT/orbsvcs/examples/Notify/ThreadPool</a></tt> +example shows how to use RTCORBA ThreadPools in RT Notification.</li> +</ul> +The performance tests for RT Notification measure the throughput, latency +and jitter for various test configurations using Periodic Suppliers: +<ul> +<li> +<tt><a href="../../orbsvcs/tests/Notify/performance-tests/scripts/1_Path_Period_0_Lanes">$TAO_ROOT/tests/Notify/performance-tests/scripts/1_Path_Period_0_Lanes</a></tt></li> + +<br>This configuration measures performance for a single data path when +the load is increased. The Supplier sends events in a single burst. +<li> +<tt><a href="../../orbsvcs/tests/Notify/performance-tests/scripts/3_Path_Period_10ms_Lanes">$TAO_ROOT/tests/Notify/performance-tests/scripts/3_Path_Period_10ms_Lanes</a></tt></li> + +<br>This configuration measures performance for 3 data paths of high, ,medium +and low priorities when the Load is increased. The Suppliers send events +at 100Hz. +<li> +<tt><a href="../../orbsvcs/tests/Notify/performance-tests/scripts/Paths_vs_Throughput">$TAO_ROOT/tests/Notify/performance-tests/scripts/Paths_vs_Throughput</a></tt></li> + +<br>This configuration has 1 high priority path ans several low priority +paths. we measure the performance of the high priority path as a function +of the number of low priority paths. +<li> +<tt><a href="../../orbsvcs/tests/Notify/performance-tests/scripts/Max_Throughput">$TAO_ROOT/tests/Notify/performance-tests/scripts/Max_Throughput</a></tt></li> + +<br>This Configuration measures the performance obtained for a data path +using a Structured Event payload: +<ul> +<li> +directly between a Supplier and Consumer in seperate processes.</li> + +<li> +directly between a colocated Supplier and Consumer.</li> + +<li> +between a Supplier and Consumer with a "Relay Consumer" as an intermidiary.</li> + +<li> +between a a Supplier and Consumer using the Notification Service.</li> +</ul> +</ul> +<b>Test Framework:</b> +<br>The <tt><a href="../../orbsvcs/tests/Notify/lib">$TAO_ROOT/orbsvcs/tests/Notify/lib</a></tt> +contains a scripting based test framework which can be used to quickly +create test cases. The framemork and its options are described <a href="notify.html#test_framework">here. </a><!-- BEGIN: CORBA Security Service Release Notes --> +<hr WIDTH="100%"> +<h3> +<a NAME="security"></a>CORBA Security Service</h3> +Point of contact: <a href="mailto:ossama@dre.vanderbilt.edu">Ossama Othman</a> +<p>Additional information is available +<a href="../Security/index.html">here</a>. + +<h4>Security Service Overview</h4> + +TAO's current Security Service support implements the core +functionality of the <a href="http://www.omg.org/cgi-bin/apps/doc?formal/02-03-11.pdf">CORBA Security Service v1.8</a>. +<p> +In particular, it implements the most of the CORBA <a href="../../orbsvcs/orbsvcs/SecurityLevel1.idl">SecurityLevel1 API</a>, +in addition to some of <a + href="../../orbsvcs/orbsvcs/SecurityLevel2.idl">SecurityLevel2</a>. +Of the transport protocols described in the above specification, only + SSLIOP is supported. Documentation for TAO's SSLIOP pluggable + transport is available <a href="../Security/index.html">here</a>. + +<p> +The 1.8 specification defines the Common Secure Interoperability version +1 (CSIv1) protocol that is currently implemented in TAO. + +<p> +There are basically two ways to use security in TAO: +<ol> +<li> +Use TAO's SSLIOP pluggable transport in TAO alone. This allows one +to secure application requests without modifying the application code. +This is the easiest approach but is also the least flexible. +</li> +<li> +Use the Security Service API implemented by TAO in conjunction with +TAO's SSLIOP pluggable transport. This provides the benefits of +secured application requests with the flexibility of disabling +security in some requests, if so desired. This approach also allows +one to choose at run-time which X.509 certificates will be associated +with application requests, as opposed to setting configuring only one +SSL certificate at application start-up-time. These things are +basically configured using the SecurityLevel2 or SecurityLevel3 +defined policies: + +<blockquote><code> + SecurityLevel2::QOPPolicy<br> + SecurityLevel2::EstablishTrustPolicy<br> + SecurityLevel3::ContextEstablishmentPolicy<br> +</blockquote></code> +</li> +</ol> + +<h4>Implemented Features</h4> +IIOP over SSL integration via TAO's <a href="../Security/SSLIOP.html">SSLIOP +pluggable transport</a>. +<ul> +<li> +Added an <tt>SSLIOP::Current</tt> implementation that can be used to obtain +the SSL session state for the current execution context. This is useful +for obtaining the SSL peer certificate chain associated with the current +request, for example.</li> + +<li> +TAO's SSLIOP pluggable transport now registers a secure invocation server +request interceptor. It enforces secure invocation by rejecting requests +coming in on the insecure port if the server is configured to do so (default +behavior).</li> + +<li> +Implemented <tt>SecurityLevel1</tt> for the SSLIOP security mechanism.</li> + +<li> +The <tt>SecurityLevel2::QOPPolicy</tt> policy has been implemented. This +policy is used to set the desired invocation Quality-of-Protection (QoP). +It can be created using <tt>ORB::create_policy()</tt>, and used in conjunction +with the standard policy manipulation CORBA features (e.g. <tt>PolicyManager</tt>, +<tt>PolicyCurrent</tt>), +meaning that this policy can be set on a per-ORB, per-thread or per-object +basis.</li> + +<p>This policy makes it possible to, for example, make both secure and +insecure invocations within the same client process. +<li> +TAO's SSLIOP pluggable transport implementation is now thread-safe.</li> + +<li> +The <tt>SecurityLevel2::EstablishTrustPolicy</tt> policy has been implemented. +This policy is used to set the desired invocation level of establishment +of trust. It can be created using <tt>ORB::create_policy()</tt>, and used +in conjunction with the standard policy manipulation CORBA features (e.g. +<tt>PolicyManager</tt>, +<tt>PolicyCurrent</tt>), meaning that this policy can be set on a per-ORB, +per-thread or per-object basis.</li> + +<p>This policy makes it possible to, for example, make authenticated and +non-authenticated invocations within the same client process. +<li> +Implemented <tt>SecurityLevel2::PrincipalAuthenticator</tt> support for +SSLIOP. In particular, a SSLIOP-specific <tt>SecurityReplaceable::Vault</tt> +implementation is now available.</li> +<li> + Implemented basic (stateless) CSIv2 support. Advanced CSIv2 + features, such as identity assertion, are currently under + development. + +</ul> +</ul> +<h4>Current Status</h4> +<ul> +<li> + Began implementation of the interfaces in the + <tt>SecurityReplaceable</tt> IDL module. They provide the + ability to replace key security components in the ORB with + another implementation with ease. Thus, the security + components in the ORB become highly extensible.</li> +<li> + Development of core <tt>SecurityLevel2</tt> interfaces such as + <tt>Credentials</tt>, <tt>SecurityManager</tt>, + and <tt>PrincipalAuthenticator</tt> has been halted in favor + of Adiron's <tt>SecurityLevel3</tt> interfaces</li> +<li> + Advanced CSIv2 features, such as identity assertion, are + currently under testing.</li> +</ul> +<h4>Schedule</h4> +<ul> +<li> +<font color="#0000FF">August 2004</font></li> + +<ul> +<li> + Complete integration of the Common Secure Interoperability + version 2 (CSIv2) protocol defined in the <a + href="http://www.omg.org/cgi-bin/doc?formal/04-03-12">CORBA + 3.0.x</a> core specification, and the <a + href="http://www.omg.org/technology/documents/formal/atlas.htm">Authorization Token Layer Acquisition Service (ATLAS)</a>. + <p> + CSIv2 and ATLAS address many of the shortcomings of the + current CSIv1-based Security Service. Overviews of each are + available on the <a href="http://www.omg.org/technology/documents/formal/omg_security.htm">OMG Security</a> web page. + <blockquote><em> + Note that CSIv2 as defined by the CORBA 3.0.x specification + has no API so TAO's implementation will use a slightly + <a + href="http://adiron.com/ORBAsec/3.5.0/docs/SL3/#SecurityLevel3API">modified API</a> currently used by several ORBs defined by Adiron LLC. + </em></blockquote> + +</li> +</ul> +<!-- + <LI> + <FONT COLOR=BLUE>November 2002</FONT> + <UL> + <LI> + Request authentication via + <A HREF="http://web.mit.edu/kerberos/www/">Kerberos</A>, as + detailed in the CORBA Security Service specification. This + requires the Common Secure Interoperability protocol (CSIv2), + but initial experiments will be performed via TAO's existing + SSLIOP pluggable transport. + </LI> + <LI> + Partial implementation of the <CODE>SecurityLevel2</CODE> + interfaces. This includes the <CODE>Credential</CODE> and + <CODE>Current</CODE> objects. + </LI> + <LI> + Partial implementation of the <CODE>SecurityAdmin</CODE> + interfaces. + </LI> + <LI> + Begin implementation of <CODE>SecurityAudit</CODE> objects. + As a side benefit, this project will allow us to research how + to secure existing services such as + <A HREF="ec.html">TAO's Real-Time Event Service</A>. + </LI> + </UL + </LI> +<li> +<font color="#0000FF">December 2002</font></li> + +<ul> +<li> +Begin implementation of the <i>Resource Access Decision (RAD) Facility</i> +specification for TAO.</li> +</ul> + +<li> +<font color="#0000FF">February 2003</font></li> + +<ul> +<li> +Begin implementation of the <i>Security Domain Membership Management (SDMM)</i> +specification for TAO. It is needed to support security domains.</li> +</ul> + +<li> +<font color="#0000FF">April 2003</font></li> + +<ul> +<li> +SSLIOP and SECIOP performance enhancements.</li> +</ul> +--> +</ul> +<!-- END: CORBA Security Service Release Notes --> + +<!--#include virtual="ec.html" --> +<p> +<hr> +<h3> +<a NAME="scheduling"></a>TAO's Scheduling Service</h3> +Point of contact: <a href="mailto:cdgill@cs.wustl.edu">Chris Gill</a> and +<a href="mailto:levine@cs.wustl.edu">David +Levine</a> +<p>Currently Implemented Features: +<ul> +<li> +The scheduling service can be built to use either a null implementation +or a strategized implementation of the configuration scheduler.</li> + +<li> +The null scheduler implementation, which is built by default, allows the +configuration scheduler to be used with applications that require a scheduling +service interface, but do not (at least in the current stage of their development, +in certain configurations, etc.) make use of the real-time scheduling features +it provides.</li> + +<li> +The strategized scheduler implementation can be built by #defining TAO_USES_STRATEGY_SCHEDULER, +and the appropriate scheduling strategy macro (TAO_USES_RMS_SCHEDULING, +TAO_USES_EDF_SCHEDULING, TAO_USES_MUF_SCHEDULING, or TAO_USES_MUF_SCHEDULING) +in $ACE_ROOT/ace/config.h. This allows the configuration scheduler to be +used with applications that require a specific scheduling strategy. Each +scheduling strategy will produce a set of static scheduling priorities, +which it will assign to operations based on their RT_Infos. For each static +priority, a strategy will also determine the run-time (dynamic) scheduling +strategy to use for that priority level.</li> +</ul> +Future work: +<ul> +<li> +Implement heap-based dispatching queues.</li> + +<li> +Add support for additional configurability, especially in the type of dispatching +strategy (list vs. heap) that will be used to dispatch operations at a +given static priority level.</li> + +<li> +Benchmark the various alternative strategies to obtain performance profiles +across different operation loads and OS platforms.</li> + +<li> +Add increased functionality. Requests and suggestions are welcome.</li> +</ul> + +<hr> +<h3> +<a NAME="fault_tolerance"></a>TAO's support for FT services</h3> +Point of contact: <a href="mailto:bala@dre.vanderbilt.edu">Balachandran +Natarajan</a> +<h4> +Current Status:</h4> +TAO supports the ORB level requirements to achieve Fault Tolerance for +CORBA Objects. The details of the ORB level support is described in the +FT chapter of the +<a href="http://www.omg.org/cgi-bin/doc?formal/02-06-01">CORBA +3.0 specification</a>. Specifically TAO implements the sections 23.2.2, +23.2.3, 23.2.6 thru 23.2.8 of the document. + +<p> +Basic support for <a href="ftcorba_services.html">FT CORBA services</a> has been added. +</p> + +<h4> +Future Work:</h4> + +<ul> +<li> +Implement 23.2.9 of the document.</li> +</ul> + +<hr> +<h3> +<a NAME="loadbalancer"></a>Load Balancer</h3> +Point of contact: <a href="mailto:ossama@dre.vanderbilt.edu">Ossama Othman</a> +<h4> +Current Status:</h4> +TAO's +<a href="http://www.cs.wustl.edu/~schmidt/PDF/load_balancing.pdf">Load +Balancer</a> currently implements the latest revision of the OMG <i>Load +Balancing and Monitoring</i> proposed specification. +<p>It provides many features and advantages over the previous prototype. +Those features and advantages include: +<ul> +<li> +Multiple object group support</li> + +<li> +Extensible load balancing strategies through IDL interfaces</li> + +<li> +Extensible load monitoring</li> + +<li> +Both "push" and "pull" style monitoring are supported</li> + +<li> +Support for infrastructure and application controlled object group membership</li> + +<li> +Improved server-side transparency</li> +</ul> +The current proposed <i>Load Balancing and Monitoring</i> specification +defines three built-in load balancing strategies. They are: +<ol> +<li> +<tt>RoundRobin</tt> <font color="#0000FF">(non-adaptive)</font></li> + +<li> +<tt>Random</tt> <font color="#0000FF">(non-adaptive)</font></li> + +<li> +<tt>LeastLoaded</tt> <font color="#0000FF">(adaptive)</font></li> + +</ol> +TAO implements all of these and the following load balancing strategies: +<ol> +<li> +<tt>LoadAverage</tt> <font color="#0000FF">(adaptive)</font></li> + +<li> +<tt>LoadMinimum</tt> <font color="#0000FF">(adaptive)</font></li> + +</ol> +<h4> +Known Issues:</h4> + +<ul> +<li> +CPU load monitoring is not working in Windows. +</li> +</ul> + +<h4> +Recent Work:</h4> + +<ul> +<li> +Implemented <tt>LoadAverage</tt> load balancing strategy.</li> + +<li> +Implemented <tt>LoadMinimum</tt> load balancing strategy.</li> + +<li> +Implemented a CPU load utilization monitor in Linux. +</li> +</ul> + +<h4> +Future Work:</h4> + +<ul> +<li> +Implement cooperative (distributed) load balancing support</li> + +<li> +Integrate multicast support</li> +<li> +Implement a middleware load balancing performance measurement toolkit</li> +<li> +Implement self adaptive load balancing strategies</li> + +</ul> + +<hr> +<h3> +<a NAME="miop"></a>Multicast InterORB Protocol (MIOP)</h3> +Point of contact: <a href="mailto:fhunleth@cs.wustl.edu">Frank Hunleth</a> +<h4> +Current Status:</h4> +The final <a href="http://www.omg.org/cgi-bin/doc?ptc/2001-11-08">MIOP +specification</a> has recently been adopted by the OMG. TAO's MIOP support +(located in $TAO_ROOT/orbsvcs/orbsvcs/PortableGroup) enables servants to +receive requests sent to multicast addresses. This is performed by creating +a GroupId that identifies the multicast group and associating it with one +or more servants. Additionally, the Unreliable IP Multicast (UIPMC) pluggable +protocol is used to send and receive multicast requests. Multicast endpoints +can be created dynamically at runtime. +<h4> +Known Issues:</h4> + +<ul> +<li> +MIOP packet reassembly is not performed yet, so the maximum request size +is limited to about 5-6kb depending on the platform.</li> +</ul> + +<h4> +Recent Work:</h4> + +<ul> +<li> +Initial check in of MIOP implementation.</li> +</ul> + +<h4> +Future Work:</h4> + +<ul> +<li> +Allow group references to be disassociated when no longer needed.</li> + +<li> +Implement MIOP packet reassembly.</li> + +<li> +Implement a Multicast Group Manager (MGM).</li> + +<li> +Implement a Multicast Gateway (MG).</li> +</ul> + +<hr> +<h3> +<a NAME="apps"></a>Test & Performance Tests</h3> +[Note: This section is not uptodate. Use with caution. February 2004.] +<br> +Point of contact: <a href="mailto:naga@cs.wustl.edu">Nagarajan Surendran</a> +<p>Current Status: +<p>The TAO IDL_Cubit test application makes use of the Naming Service and +the server holds a TAO_Naming_Server component.Just running server and +client is enough to test the application. +<p>The various tests in the tests/POA test the different features of the +Portable Object Adapter interface like Explicit Activation, On Demand Activation,etc.. +<p>MT_Cubit: +<p>Current status: +<p>The TAO MT_Cubit test application is meant to serve as a starting point +for real-time tests on the TAO system. It comprises the following parts: +<ul> +<li> +<i>Server.</i> The server creates multiple CORBA objects (servants), each +with different real-time priorities. This priority is implemented by using +real-time thread support provided by the operating system. Thus, requests +sent to a high-priority servant are handled by a high-priority real-time +thread, and those sent to a lower priority servant are handled by correspondingly +lower priority threads.</li> + +<li> +<i>Client.</i> The client component binds to the servants, and sends a +stream of CORBA requests to the servants. It measures the response time, +i.e. the time taken for the request to complete successfully. In particular, +it measures the time taken for requests sent to the high priority servant +to complete. The volume of lower priority requests is configurable. The +client is thus able to measure the performance of the high-priority servant +in the presence of competition from several lower-priority servants.</li> +</ul> +Clearly, if the ORB endsystem handles the priorities of the various requests +correctly, increasing the volume of lower priority requests should not +affect the performance seen by the higher priority requests. The application +thus serves as a tool to measure and confirm this behavior. +<p>Future work: +<ul> +<li> +Study the impacts of scheduling & concurrency strategies on performance.</li> + +<li> +Evolve into a testbed for discovering sources of performance non-determinism +& priority inversion.</li> +</ul> +Pluggable: +<p>Current status: +<p>The TAO Pluggable test utilizes ACE Timeprobes to time the latency at +various points in the ORB, especially that incurred by the Pluggable Protocols +implementation. Comparisons can be made not only between different layers +of the ORB, but also between different protocols as they become available. +<p>Future work: +<ul> +<li> +Add options to redirect the output to a file.</li> + +<li> +Script or otherwise automate the piping of the output to a spreadsheet.</li> +</ul> + +<hr> +<h3> +<a NAME="ace"></a>ORB-related ACE Changes</h3> +Points of contact: <a href="mailto:nanbor@cs.wustl.edu">Nanbor Wang</a> +and <a href="mailto:irfan@cs.wustl.edu">Irfan Pyrarli</a> +<p>Recently Completed Work: +<ul> +<li> +Added special declaration to OS.h for <tt>inet_ntoa</tt> and other functions +because VxWorks doesn't provide full argument prototypes for these library +functions.</li> + +<li> +The current caching connector behaves properly in the face of a non-blocking +connect request. The "fix" is simply to not support non-blocking connects +through the cache. When the <tt>connect()</tt> fails with <tt>EWOULDBLOCK</tt>, +morph the error to -1 and clean up the request.</li> + +<li> +Service handlers obtained from the caching connector are now cleaned up. +The application needs to be able to signal that it's not using it any longer, +and, when the application encounters an error, needs to effectively close +down that connection for good so that a new connection can be initiated.</li> + +<br>Added the ability for a Svc_Handler to recycle itself. idle() can be +called when the Svc_Handler is done serving a particular connection and +can how be recycled. The Svc_Handler now also has a pointer to a recycler +that is responsible for managing the connections. The recycler is usually +a Cached_Connector. +<br>Added new class ACE_Recycling_Strategy. It defines the interface (and +default implementation) for specifying a recycling strategy for a Svc_Handler. +This strategy acts as a consular to the Svc_Handler, preparing it for the +tough times ahead when the Svc_Handler will be recycled. +<br>Added new class ACE_NOOP_Concurrency_Strategy. It implements a no-op +activation strategy in order to avoid calling open on a recycled svc_handler +multiple times. +<br>ACE_Cached_Connect_Strategy now implements the ACE_Connection_Recycling_Strategy +interface. This allows Svc_Handlers to cache themselves with ACE_Cached_Connect_Strategy +when they become idle. It also allows them to purge themselves from the +connection cache when the Svc_Handlers close down. +<br>Also added ~ACE_Cached_Connect_Strategy that will cleanup up the connection +cache.</ul> +Future work: +<blockquote><i>None currently scheduled.</i></blockquote> + +<hr> +<h3> +<a NAME="dove"></a>The DOVE Demo</h3> +Points of contact: <a href="mailto:mk1@cs.wustl.edu">Michael Kircher</a> +and <a href="mailto:cdgill@cs.wustl.edu">Chris Gill</a>. +<p><a href="http://www.cs.wustl.edu/~schmidt/dove.html">DOVE</a> is documented +in detail <a href="http://www.cs.wustl.edu/~schmidt/Dove.ps.gz">online</a>. +This discussion focuses on the following goals: +<ul> +<li> +Have a DOVE Browser running using Java Beans as vizualization components.</li> + +<li> +Have the Event Channel as DOVE Agent running with an Event Consumer in +the DOVE Browser.</li> + +<li> +Having a DOVE Management Information Base (MIB), which dumps all events +transfered on the Event Channel into a file on persistent storage for later +reuse.</li> +</ul> +The DOVE Browser uses independent visualization components (Java Beans) +and the Event Channel as DOVE Agent. Connections can be established between +monitored metrics and the visualization components. +<p>We have three major components: Observables (monitored metrics), Observers +(a Java Bean for displaying the metric) and a DataHandler (for demultiplexing +the monitored metrics to the appropriate Observables). Each component inherits +from a base class, so that a certain behavior of the components can be +assured for each component. Relationships between components are based +on these base classes. +<p>The used Java Beans are required to conform to some standards, as they +have to support a function called "getProperty" which allows the DOVE Browser +to determine if the vizualization capabilities of a specific Java Bean +are sufficient to display the metric. A JavaBean is for example a Java +Panel which shows a Graph of the delivered doubles. So all metrics can +be displayed by this visualization component which can be expressed by +a single double. +<p>The DataHandler is connected to the Event Push Consumer (PUSH, because +we use the push concept of the Event Service). The Event Push Consumer +does not know what kind of data is transported. The only component knowing +all the details about the dependencies of the metrics is the DataHandler. +This separation allows easy extension and change of the demo. +<p><a href="http://students.cec.wustl.edu/~mk1/dove.html">Object Diagrams</a> +are available about this new concept. +<p>Event Service events are used as communication between DOVE Applications +and the DOVE Browser. The DOVE MIB analyses the event data field of all +events and stores this information into a file. The event data filed is +of type CORBA::Any and the DOVE MIB has no notion of what is conveyed in +this field. So the DOVE MIB has to discover the content via the embedded +type code information. Future work includes: +<ul> +<li> +Enhancing MIB functionality</li> + +<li> +Monitoring the AV Streaming Service</li> +</ul> +For more information on the DOVE demo, please refer to: $TAO_ROOT/orbsvcs/tests/Simulator/README. +<p> +<hr> +<h3> +<a NAME="forwarding"></a>Location Forwarding</h3> +Point of contact: <a href="mailto:irfan@cs.wustl.edu">Irfan Pyarali</a>, +<a href="mailto:mk1@mk1.wustl.edu">Michael +Kircher</a>. +<p>For more information see <a href="../forwarding.html">Location forwarding</a> +<p> +<hr> +<h3> +<a NAME="leader"></a>Global Resources and Leader-Follower Model</h3> +Point of contact: <a href="mailto:irfan@cs.wustl.edu">Irfan Pyarali</a>, +<a href="mailto:mk1@mk1.wustl.edu">Michael +Kircher</a>. +<p>For more information see <a href="../leader_follower.html">Leader-follower +model</a> +<p> +<hr> +<h3> +<a NAME="locate"></a>Implementation of locate request</h3> +Point of contact: <a href="mailto:irfan@cs.wustl.edu">Irfan Pyarali</a>, +<a href="mailto:mk1@mk1.wustl.edu">Michael +Kircher</a>. +<p>For more information see <a href="../locate_request.html">Locate request</a> +<p> +<hr> +<h3> +<a NAME="ami"></a>Asynchronous Method Invocation</h3> +Points of contact: <a href="mailto:alex@cs.wustl.edu">Alexander Arulanthu</a> +, <a href="mailto:Michael.Kircher@mchp.siemens.de">Michael Kircher</a> +and +<a href="mailto:coryan@uci.edu">Carlos O'Ryan</a> +<p>Status: +<ul>We've implemented the callback model of the +<a href="http://www.omg.org/cgi-bin/doc?formal/01-12-60">CORBA +Messaging specification</a>. To activate the AMI for TAO and the TAO IDL +compiler define <tt>TAO_HAS_CORBA_MESSAGING</tt>, +<tt>TAO_HAS_AMI_CALLBACK</tt> +in your config.h file. The TAO IDL compiler can generate the AMI "callback" +stubs, ReplyHandler und reply stubs using the <tt>-GC</tt> switch. +<p>For an example see <tt>$TAO_ROOT/tests/AMI</tt> and <tt>$TAO_ROOT/examples/AMI</tt>.</ul> +Finished work: +<ul> +<li> +Redesign of the IDL compiler to make an addtional pass over the AbstractSyntaxTree +and generate the implied-IDL code in memory. This reduced the amount of +AMI specific IDL compiler code dramatically.</li> + +<li> +Support for exceptions</li> + +<li> +Support for attributes</li> + +<li> +Support for buffering and batching AMI calls. See <a href="../../examples/Buffered_AMI">Buffered +AMI</a> example for details.</li> + +<li> +Support for deferred synchronous invocations. +<a href="mailto:j.parsons@vanderbilt.edu">Jeff +Parsons</a></li> + +<li> +Support for timeouts in combination with AMI calls, response handler gets +CORBA::TIMEOUT exception on timeout</li> +<li> +The AMI support in TAO 1.4.7 is as described in the 2.6 spec, from 1.4.8 we support +by default the 3.0.3 described mapping.</li> +</ul> +Future Work: +<ul> +<li> +Implementation of the poller model.</li> +</ul> +<hr> +<h3> +<a NAME="csd"></a>Custom Servant Dispatching</h3> +Points of contact: <a href="mailto:bradley_t@ociweb.com">Tim Bradley</a> +<h4> +Current Status:</h4> +<ul>This Custom Servant Dispatching (CSD) feature provides user applications with the ability to implement and "plug-in" custom strategies to handle the dispatching of requests to servants. +<p>A concrete CSD Strategy implementation has also been added to serve as a "reference implementation". This is being called the CSD Thread Pool Strategy (TP_Strategy). The TP_Strategy provides a means to decouple the threads (ORB threads) that receive requests from the underlying transport from the thread that will ultimately dispatch the request to the target servant object. The TP_Strategy implements a "request queue" as the integral part of the mechanism that allows an ORB thread to "hand-off" a request to one of the TP_Strategy object's worker threads. The TP_Strategy reference implementation is provided as an example of how concrete CSD Strategy could be implemented. +<p>Two approaches are supported for applying CSD strategy to an application. +<ol> +<li>Explicitly calling CSD interfaces. +<p>Here is an example application code showing how a TP_Strategy object can be created and applied to a POA: + +<pre> + + PortableServer::POA_var poa = ...; // create the poa. + + // Create a new TP_Strategy object and save it into a "smart pointer" variable. + TAO::CSD::TP_Strategy_Handle csd_strategy = new TAO::CSD::TP_Strategy(); + + // Set the number of threads before calling apply_to(). + csd_strategy->set_num_threads(2); + + // Set the servant serialization flag before calling apply_to(). + csd_strategy->set_servant_serialization (false); + + // Tell the strategy to apply itself to the poa. + if (csd_strategy->apply_to(poa.in()) == false) + { + ACE_ERROR((LM_ERROR, "Failed to apply CSD strategy to the poa.\n")); + return -1; + } +</pre> +<li>Service Configurator + + <p>The format of the CSD specific parameters for creating the TP_Strategy service object is: + <pre>-CSDtp <poa_name>:<csd_thread_number>:[OFF]</pre> + + <p>The last portion of the parameter is the servant serialization flag. It's only needed when the servant serialization needs be turned off, otherwise the servant serialization is always on. When servant serialization is on (the default), the TP_Strategy will serialize requests to any particular servant. Requests to different servant objects can occur in parallel, but requests to any particular servant will be dispatched serially (ie, one at a time). + + <p>Here is an example of the svc.conf file. + + <pre> + dynamic TAO_CSD_TP_Strategy_Factory Service_Object * + TAO_CSD_ThreadPool:_make_TAO_CSD_TP_Strategy_Factory() "-CSDtp RootPOA:2" + </pre> +</ol> +</ul> +<h4> +Known Issues:</h4> +<ul> +<li> +This feature is not currently tested for VxWorks. +</li> +<li> +This feature does not support the operation of the various server side 'Current' implementations (e.g. PICurrent, POACurrent, etc.). It should not be used by applications that require these to function correctly. See issues <a href="http://bugzilla.dre.vanderbilt.edu/show_bug.cgi?id=3087">3087</a> and <a href="http://bugzilla.dre.vanderbilt.edu/show_bug.cgi?id=3327">3327</a>. +</li> +</ul> +<hr> +<h3> +<a NAME="interceptor"></a>Portable Interceptors</h3> +Point of contact: <a href="mailto:ossama@dre.vanderbilt.edu">Ossama Othman</a>. +<p>For more information see <a href="../interceptors.html">Portable Interceptors</a> +<br> +<hr> +<h3> +<a NAME="localobject"></a>Local Interfaces</h3> +Point of contact: <a href="mailto:nanbor@cs.wustl.edu">Nanbor Wang</a>. +<p>Local interfaces are first defined in the CORBA Component Model specification. +For more information on using the local interfaces, please refers to Section +11.1.1 to 11.1.4 of the +<a href="http://www.omg.org/cgi-bin/doc?orbos/99-07-01/">spec</a> +and our short <a href="../LocalObject.html">guideline</a> on implementing +local objects. +<br> + + +<hr> +<h3> +<a NAME="sciop"></a>SCIOP Support in TAO</h3> +Point of contact: <a href="mailto:gthaker@atl.lmco.com">Gautam H. Thaker</a>, Lockheed Martin Advanced Technology Labs, Cherry Hill, NJ. + + +<p>TAO has support for OMG's GIOP SCTP Protocol mapping <a +href=http://www.omg.org/cgi-bin/doc?ptc/2003-08-20>spec</a>, except +that SCIOP protocol properties are not yet supported. Extensive +information about SCTP and how it may be used from both ACE and TAO is +available in several README files in $ACE_ROOT/performance-tests/SCTP/ +directory. ACE+TAO's SCTP support can be used either with Linux's +OpenSS7 SCTP implementation or with Linux's LKSCTP SCTP +implementation. <p> A paper describing the ACE+TAO SCTP +implementation and measured results are available <a +href=http://www.atl.external.lmco.com/projects/QoS/documents/DOA2003_97_Thaker.pdf> +online</a>.<P> + +<hr> +<h3> +<a NAME="ipv6"></a>IPv6 Support in TAO</h3> +Point of contact: <a href="mailto:mcorino@remedy.nl">Martin Corino</a>, Remedy IT. + + +<p>TAO has support for IPv6 for IIOP under Windows and Linux. +To use this, add <code>ipv6</code> to your <code>default.features</code> +file and regenerate all makefiles. When using the automated tests, add IPV6 +to the configs. +<p> +Finished work: +<li>Added IPv6 support to all IIOP related classes (parsers, connectors, + acceptors)</li> +<li>Added IPv6 support to corbaloc and mcast URL parsers.</li> +<li>Implemented IPv6 support in the TAO_IOR_MCast utility class from TAO Svc + Utils</li> +<p> + +The following gotchas are known: +<li>In a localhost situation connecting a server listening at the IPv6 ANY address + has the following problems: +On Linux this does not work when the client tries to connect to a + LinkLocal address of one the local NICs, +on Windows this only works (locally) by using the 'localhost' address (either + IPv4 or IPv6) +</li> +<li>Usage of the '-ORBObjRefStyle url' switch poses problems in localhost situations + on Windows since this switch causes 'object_to_string' to only return the first server + endpoint from the IOR profile as a corbaloc URL. In IPv6 servers this will (almost) allways be an IPv4 interface address of one + of the local NICs. This will than cause problems when the server is listening at + the IPv6 ANY address as described above. +</li> + +<p> +Future work: +<li>IPv6 support for other protocols/strategies than IIOP (i.e. SHMIOP, UIOP, + SCIOP, SSLIOP etc.; for this we excluded some regression tests in IPv6 builds for this + reason)</li> + +<li>IPv4 runtime dependencies in various CORBA services (f.i. Event service) +</li> + + +<P> + + + +<hr> +<h3> +<a NAME="CORBA-conformance"></a>CORBA Standards Conformance</h3> +Here is a summary of TAO's conformance issues with CORBA latest CORBA specifications +(updated 9 August 2000): +<br>2.3.1 and 2.3 differ in very little, if at all, check: +<br><a href="ftp://ftp.omg.org/pub/docs/formal/99-10-07.pdf">ftp://ftp.omg.org/pub/docs/formal/99-10-07.pdf</a> +<br>and search for the change bars, meanwhile this can help: +<ul> +<li> +The full OBV (valueboxes, valuetypes in all sorts of things, value graphs, +Anys and value types, abstract interfaces, etc. etc.)</li> + +<li> +The IfR database cannot be populated (so it is less than fully operational +;-))</li> + +<li> +Fixed data types (who cares)</li> + +<li> +GIOP fragments (1.1 and 1.2) are not completely tested.</li> + +<li> +The old ServantManager interfaces (POA stuff)</li> + +<li> +Domain Managers (useful for security, but otherwise nobody seems to care)</li> + +<li> +I think our interpretation of codesets is compliant (we only support one +codeset), but would have to check.</li> +</ul> +Future Work (aka. known problems): +<ul> +<li> +Supposedly, any constructed types that contains local types become local +automatically. TAO_IDL currently doesn't handle the array type very well +if one is defined outside the scope of a local interface.</li> + +<li> +Need to test local object support more systematically and comprehensively. +(Does TAO throw a MARSHAL exception when trying to marshal a local type?)</li> +</ul> + +<hr> +<p>Back to the TAO <a href="../index.html">documentation index</a>.<!--#include virtual="/~schmidt/cgi-sig.html" --> +<hr><!-- hhmts start -->Last modified: Thu May 4 13:08:35 UTC 2006<!-- hhmts end --> +</body> +</html> |