SlideShare a Scribd company logo

20 Tips for OpenSplice Newbies

Angelo Corsaro, profile picture
Angelo Corsaro

The document provides 20 tips for beginners using OpenSplice, focusing on the fundamentals of domains, partitions, and topics in DDS applications. Key insights include how to choose a domain, configure shared memory, and manage Quality of Service (QoS) settings effectively. Additionally, it emphasizes the importance of understanding lifecycle management and data access methods to enhance application performance.

TechnologyBusiness
1 of 53
Downloaded 305 times
1
2
3
Most read
4
5
6
7
8
9
10
11
Most read
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Most read
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
20 Tips for OpenSplice Newbies Angelo Corsaro, Ph.D. Chief Technology Officer OMG DDS Sig Co-Chair PrismTech angelo.corsaro@prismtech.com
Domain
Tip #0 Domains, Partitions, Topics Domain (e.g. Domain 123) ¨ All DDS applications publish Copyright  2010,  PrismTech  –    All  Rights  Reserved. and subscribe data on a Partition belonging to a given domain Partitions (e.g. Partition “Telemetry”) ¨ Partitions are defined by means for strings and can Topic Instances/ be matched with regular Samples expressions ¨ If not explicitly specified the default partition in the default domain is automatically chosen
Tip #0 Partitions Matching "building-1.floor-3.room-51" Copyright  2010,  PrismTech  –    All  Rights  Reserved. "building-1.floor-1.room-*" ... building-1.floor-3.room-5 ... ¨ Partitions are defined by means building-1.floor-1.room-111 building-1.floor15-.room-51 for strings and can be matched with building-1.floor-3.room-11 regular building-1.floor-1.room-1 building-1.floor-10.room-100 expressions] ... "building-1.floor-*.room-11?" Domain
Configuration 101
Tip #1 Choosing a Domain ¨ In OpenSplice DDS v5.x a domain is selected by: Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ Defining the OSPL_URI environment variable ¨ Passing a URI pointing at the the domain XML configuration file at OpenSplice startup ¨ Passing the URI of the configuration file as a string parameter of the DomainParticipantFactory::create_participant method ¨ Passing the name of the domain as specified in the configuration file to the DomainParticipantFactory::create_participant method ¨ Passing the empty string to represent the default domain to the DomainParticipantFactory::create_participant method
Tip #1 Choosing a Domain Domain specified via URI Copyright  2010,  PrismTech  –    All  Rights  Reserved. Defining Domain Configuration at Startup dpf.create_participant( “file:///some/path/myospl.xml”, $ ospl start file:///some/path/myospl.xml qos, listener, mask); Default Domain dpf.create_participant( Domain specified via a Domain Name “”, qos, dpf.create_participant( listener, “MyDomainNameAsSpecifiedOnTheXMLFile”, mask); qos, listener, mask);
Tip #1 Domains on OpenSplice v6.x ¨ The DDS specification did not originally define the type for the Copyright  2010,  PrismTech  –    All  Rights  Reserved. DomainId, as a result vendors where free to choose their on types ¨ As the DDSI/RTPS specification defines a DomainId with and integer it makes sense to uniform the DDS API to use an integer DomainID ¨ As a result, starting with OpenSplice DDS the domain will be selected specifying its associated id: dpf.create_participant( 15, qos, listener, mask);
Tip #2 Start OpenSplice First Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ OpenSplice v5.x runs by default on a shared/ memory + daemon configuration ¨ As such, if you forget to start the infrastructure your application will fail at start up ¨ Thus always recall to run: $ ospl start
Tip #3 Shared Memory Size OpenSplice DDS shared memory size is defined in its Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ configuration file. The size defined by the default configuration file is 10MBytes ¨ Beware that different OS have different limitations w.r.t. the maximum shared memory segment that can be allocated ¨ If you want to go beyond the OS limits you need to change the configuration of your kernel
Tip #3 Linux ¨ The default value for the maximum shared memory segment is 32MBytes Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ This default can be changed in several ways (1) Adding this line to your /etc/rc.d/rc.local file: echo “your_max_shared_memory_size” > /proc/sys/kernel/shmmax (2) Changing the settings for the sys-limits (save changes on /etc/ sysctl.conf to maintain them across reboots): $ sysctl -w kernel.shmmax=yourMaxValue
Tip #3 Windows The default maximum size for Shared Memory Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ segments on Windows is 2GB ¨ To exend it, say to 3GB, add the /3GB the boot.ini as shown below: [boot loader] timeout=30 default=multi(0)disk(0)rdisk(0)partition(1)WINDOWS [operating systems] multi(0)disk(0)rdisk(0)partition (1)WINDOWS="Windows NT Workstation Version 4.00" /3GB
DDS Topics
Tip #4 Topic Types & Keys Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ Topic Types can define some of their attributes as keys ¨ Yet, even when a Topic type does not define a key the keylist directive has to be provided -- just to tell the IDL compiler that this is a topic
Tip #4 Topic Types & Keys Keyful Keyless Copyright  2010,  PrismTech  –    All  Rights  Reserved. enum TemperatureScale { CELSIUS, FAHRENHEIT, struct UniversalConstants { KELVIN double PI; }; double e; // Nepero Number double g; // Gravitational Constant struct TempSensorType { double NA; // Avogadro Number short id; double F; // Faraday Constant float temp; double K; // Boltzman Constant float hum; double c; // Light Speed TemperatureScale scale; }; }; #pragma keylist UniversalConstants #pragma keylist TempSensorType id
Tip #5 Topic Instances Copyright  2010,  PrismTech  –    All  Rights  Reserved. enum TemperatureScale { CELSIUS, Writer Reader FAHRENHEIT, KELVIN }; 2 26.0 70.0 CELSIUS 1 26.0 70.0 CELSIUS struct TempSensorType { 1 26.0 70.0 CELSIUS 2 26.0 70.0 CELSIUS short id; float temp; float hum; TemperatureScale scale; }; #pragma keylist TempSensorType
Tip #5 Topic Instances Copyright  2010,  PrismTech  –    All  Rights  Reserved. enum TemperatureScale { CELSIUS, Writer Reader FAHRENHEIT, KELVIN }; 2 26.0 70.0 CELSIUS struct TempSensorType { 1 26.0 70.0 CELSIUS 2 26.0 70.0 CELSIUS 1 26.0 70.0 CELSIUS short id; float temp; float hum; TemperatureScale scale; }; #pragma keylist TempSensorType id
Tip #6 Topic Instances Lifecycle struct VehiclePosition { string plate; Copyright  2010,  PrismTech  –    All  Rights  Reserved. long x; long y; }; com.myco.VPos #pragma keylist VehiclePosition plate struct VehiclePosition { string plate; long x; long y; }; #pragma keylist VehiclePosition com.myco.VPos
Tip #6 Topic Instances Lifecycle struct VehiclePosition { string plate; New Copyright  2010,  PrismTech  –    All  Rights  Reserved. long x; “A01” 100 200 long y; }; com.myco.VPos #pragma keylist VehiclePosition plate struct VehiclePosition { string plate; long x; long y; New }; #pragma keylist VehiclePosition “A01” 100 200 com.myco.VPos
Tip #6 Topic Instances Lifecycle struct VehiclePosition { string plate; New Copyright  2010,  PrismTech  –    All  Rights  Reserved. long x; “A01” 100 200 long y; }; New #pragma keylist VehiclePosition plate “B41” 57 31 com.myco.VPos struct VehiclePosition { string plate; long x; long y; }; #pragma keylist VehiclePosition “B41” 57 31 “A01” 100 200 com.myco.VPos
Tip #6 Topic Instances Lifecycle struct VehiclePosition { string plate; New Copyright  2010,  PrismTech  –    All  Rights  Reserved. long x; “A01” 110 210 “A01” 100 200 long y; }; New #pragma keylist VehiclePosition plate “B41” 57 31 com.myco.VPos struct VehiclePosition { string plate; long x; long y; }; #pragma keylist VehiclePosition “A01” 110 210 “B41” 57 31 “A01” 100 200 com.myco.VPos
Tip #6 Topic Instances Lifecycle New Copyright  2010,  PrismTech  –    All  Rights  Reserved. “A01” 120 220 “A01” 110 210 “A01” 100 200 New “B41” 57 31 com.myco.VPos “A01” 120 220 “A01” 110 210 “B41” 57 31 “A01” 100 200 com.myco.VPos
Tip #6 Topic Instances Lifecycle New Copyright  2010,  PrismTech  –    All  Rights  Reserved. “A01” 120 220 “A01” 110 210 “A01” 100 200 New “B41” 47 19 “B41” 57 31 com.myco.VPos “B41” 47 19 “A01” 120 220 “A01” 110 210 “B41” 57 31 “A01” 100 200 com.myco.VPos
Tip #6 Topic Instances Lifecycle New Copyright  2010,  PrismTech  –    All  Rights  Reserved. “A01” 120 220 “A01” 110 210 “A01” 100 200 New “B41” Disposed - - “B41” 47 19 “B41” 57 31 com.myco.VPos “B41” 47 19 “A01” 120 220 “A01” 110 210 “B41” 57 31 “A01” 100 200 com.myco.VPos
Tip #7 Default Lifecycle Settings The WriterDataLifecycle controls when instances are Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ disposed. By default DDS disposes unregistered instances ¨ Automatically disposing an instance perhaps is not what your want to do when terminating your application, as this would remove persistent data for the given instance!
QoS
Tip #8 Understand the QoS Model ¨ DDS defines 22 QoS DURABILITY LIVELINESS DEST. ORDER TIME-BASED FILTER policies that can be HISTORY OWENERSHIP PARTITION RESOURCE LIMITS Copyright  2010,  PrismTech  –    All  Rights  Reserved. LIFESPAN OWN. STRENGTH PRESENTATION applied to RELIABILITY DW LIFECYCLE communication USER DATA DEADLINE DR LIFECYCLE entities to control their TOPIC DATA GROUP DATA LATENCY BUDGET TRANSPORT PRIO ENTITY FACTORY local as well as end- to-end behaviour RxO QoS Local QoS ¨ Most of the QoS Policies that control an end-to-end property follow the so-called Request vs. Offered (RxO) Model based on which the QoS requested by the Consumer should not exceed the QoS Provided by the Producer
Tip #9 History + Reliability Interplay ¨ The History QoS controls the number of samples that Copyright  2010,  PrismTech  –    All  Rights  Reserved. are maintained by DDS for a given topic ¨ DDS can keep the last n samples or keep all samples up to when they are not taken by the application ¨ The History setting has an impact on the reliability of data delivery as perceived by the application. Thus beware of your settings!
Tip #9 History + Reliability Interplay struct Counter { QoS Settings int cID; Reliability = Reliable Copyright  2010,  PrismTech  –    All  Rights  Reserved. int count; }; #pragma keylist Counter cID History = KeepLast(1) History Depth = 1 History Depth = 1 (DDS Default) (DDS Default) Network 1 1 1 2 DataReader 2 1 1 2 2 3 DataWriter 3 1 2 2 2 3 3 1 Topic Topic DataReader Cache DataWriter Cache
Tip #9 History + Reliability Interplay struct Counter { QoS Settings int cID; Reliability = Reliable Copyright  2010,  PrismTech  –    All  Rights  Reserved. int count; }; #pragma keylist Counter cID History = KeepLast(1) History Depth = 1 History Depth = 1 (DDS Default) (DDS Default) Network 1 2 1 2 DataReader 2 2 2 3 DataWriter 3 1 2 3 3 1 Topic Topic DataReader Cache DataWriter Cache
Tip #9 History + Reliability Interplay struct Counter { QoS Settings int cID; Reliability = Reliable Copyright  2010,  PrismTech  –    All  Rights  Reserved. int count; }; #pragma keylist Counter cID History = KeepLast(1) History Depth = 1 History Depth = 1 (DDS Default) (DDS Default) Network 1 2 1 2 DataReader 2 3 2 3 DataWriter 3 1 3 1 Topic Topic DataReader Cache DataWriter Cache
Tip #7 #10 Define Resource Limits ¨ DDS Provides a QoS that allow to control the amount of resources used by Copyright  2010,  PrismTech  –    All  Rights  Reserved. DataReaders and DataWriters ¨ By default, DDS does not imposes any limit, with the results that if you have a buggy application or an asymmetry in your system you might end-up consuming unbounded amount of memory -- and in the OpenSplice DDS case filling the Shared Memory ¨ To avoid this problem, always set appropriate Resource Limits for your application by defining: ¨ max_samples ¨ max_instances ¨ max_samples_per_instance
Accessing the Data
Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; ¨ DataReader::read }; iterates over the available History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID sample instances ¨ Samples are not removed from the local cache as 1 1 1 2 1 3 1 4 result of a read DataReader 2 1 2 2 2 3 3 1 3 2 3 3 3 4 3 5 ¨ Read samples can be Topic read again, by accessing the cache with the Samples Read Samples not Read proper options (more DataReader Cache later)
Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; ¨ DataReader::read }; iterates over the available History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID sample instances ¨ Samples are not removed from the local cache as 1 1 1 2 1 3 1 4 result of a read DataReader 2 1 2 2 2 3 3 1 3 2 3 3 3 4 3 5 ¨ Read samples can be Topic read again, by accessing Samples not Read the cache with the Samples Read proper options (more DataReader Cache later)
Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; ¨ DataReader::read }; iterates over the available History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID sample instances ¨ Samples are not removed from the local cache as 1 1 1 2 1 3 1 4 result of a read DataReader 2 1 2 2 2 3 3 1 3 2 3 3 3 4 3 5 ¨ Read samples can be Topic read again, by accessing Samples Read Samples not Read the cache with the proper options (more DataReader Cache later)
Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; }; ¨ DataReader::take History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID iterates over the available sample instances 1 1 1 2 1 3 1 4 Taken Samples are 2 1 2 2 2 3 ¨ DataReader 3 1 3 2 3 3 3 4 3 5 removed from the Topic local cache as result Samples not Taken of a take DataReader Cache
Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; }; ¨ DataReader::take History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID iterates over the available sample instances 1 2 1 3 1 4 Taken Samples are 2 2 2 3 ¨ DataReader 3 2 3 3 3 4 3 5 removed from the Topic local cache as result Samples not Taken of a take DataReader Cache
Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; }; ¨ DataReader::take History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID iterates over the available sample instances 1 3 1 4 Taken Samples are 2 3 ¨ DataReader 3 3 3 4 3 5 removed from the Topic local cache as result Samples not Taken of a take DataReader Cache
Tip #12 Sample, Instance, View State History Depth = 2 ¨ Along with data samples, DataReaders provides state information allowing to detect relevant Copyright  2010,  PrismTech  –    All  Rights  Reserved. 1 1 transitions in the life-cycle of data as well as data writers 2 2 3 ¨ Sample State (READ | NOT_READ): Determines DataReader SampleInfo wether a sample has already been read by this DataReader or not. 1 1 1 2 ¨ Instance State (ALIVE, NOT_ALIVE, DISPOSED). 2 2 2 3 Determines wether (1) writer exist for the specific 3 1 instance, or (2) no matched writers are currently Samples available, or (3) the instance has been disposed Topic ¨ View State (NEW, NOT_NEW). Determines wether this is the first sample of a new (or re-born) instance DataReader Cache
Tip #13 Beware of Invalid Samples! ¨ For each data sample accessed via a read or take Copyright  2010,  PrismTech  –    All  Rights  Reserved. DDS provides you with a SampleInfo ¨ The SampleInfo contains meta-information about the Sample, such as timestamp, lifecycle information, etc., but most importantly tells you if the data is valid or not! ¨ Data is not valid, when the sample you are receiving notifies things like an instance being unregistered or disposed
Tip #14 Reading Only “Fresh” Data Copyright  2010,  PrismTech  –    All  Rights  Reserved. dr.read(samples, infos, LENGTH_UNLIMITED, // read all available samples NOT_READ_SAMPLE_STATE, ANY_VIEW_STATE, ALIVE_INSTANCE_STATE);
Tip #15 Reading All Data Copyright  2010,  PrismTech  –    All  Rights  Reserved. dr.read(samples, infos, LENGTH_UNLIMITED, // read all available samples ANY_SAMPLE_STATE, ANY_VIEW_STATE, ALIVE_INSTANCE_STATE);
Tip #16 Getting Everything Copyright  2010,  PrismTech  –    All  Rights  Reserved. dr.read(samples, infos, LENGTH_UNLIMITED, // read all available samples ANY_SAMPLE_STATE, ANY_VIEW_STATE, ANY_INSTANCE_STATE); NOTE: As explained on the Tip #13 in this case you might get invalid data samples, thus have to check on their validity via the SampleInfo.valid_data attribute
Tip #17 Status vs. Read Condition Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ Both a Status as well as a Read Condition can be used to wait for data to be available on a DataReader ¨ The main difference is that a ReadCondition allows to set the exact SAMPLE, VIEW and INSTANCE status for which the condition should trigger, while the StatusCondition triggers when a sample is received
Tip #17 Status vs. Read Condition Guidelines Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ If what you really care is getting a condition that triggers for the state ANY_SAMPLE_STATE, ANY_VIEW_STATE and ANY_INSTANCE_STATE than use a StatusCondition as this is more efficient than a ReadCondition ¨ If you are interested in having a condition that triggers for specific a status then use the ReadCondition
Memory Management
Tip #18 Return Memory Loans Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ The DataReader read/take allocate loan memory to the application when the length of the containers passed for storing samples and info is zero ¨ In this case the loaned memory must be returned via a return_loan operation!
Return Memory Loans Tip #18 Copyright  2010,  PrismTech  –    All  Rights  Reserved.
Tip #19 Beware of Strings Ownership ¨ The DDS C++ API takes ownership of the string you pass Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ As a result, you need to understand when it is necessary to “duplicate” a string ¨ To this end, DDS provides the DDS:string_dup call to facilitate this task subQos.partition.name.length (1); subQos.partition.name[0] = DDS::string_dup (read_partition);
Concluding Tip
Tip #20 Read, Write, Ask Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ Read the Manuals and if possible the Specification ¨ Write your own code examples ¨ Don’t be shy to ask questions on the OpenSplice mailing list
:: Connect with Us :: Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¥ opensplice.com ¥ forums.opensplice.org ¥ @acorsaro ¥ opensplice.org ¥ opensplicedds@prismtech.com ¥ @prismtech ¥ crc@prismtech.com ¥ sales@prismtech.com ¥ youtube.com/opensplicetube ¥ slideshare.net/angelo.corsaro

More Related Content

PDF
10 Reasons for Choosing OpenSplice DDS
Angelo Corsaro
 
PDF
The DDS Tutorial - Part I
Angelo Corsaro
 
PDF
OpenSplice DDS Tutorial -- Part II
Angelo Corsaro
 
PDF
Getting Started in DDS with C++ and Java
Angelo Corsaro
 
PDF
Advanced OpenSplice Programming - Part I
Angelo Corsaro
 
PPTX
Optimize S1000D & ATA Technical Illustration production
Vizualsite LLC
 
PDF
Getting Started with DDS in C++, Java and Scala
Angelo Corsaro
 
PDF
Materi SQL Query Berjenjang
Ngurah Endra
 
10 Reasons for Choosing OpenSplice DDS
Angelo Corsaro
 
The DDS Tutorial - Part I
Angelo Corsaro
 
OpenSplice DDS Tutorial -- Part II
Angelo Corsaro
 
Getting Started in DDS with C++ and Java
Angelo Corsaro
 
Advanced OpenSplice Programming - Part I
Angelo Corsaro
 
Optimize S1000D & ATA Technical Illustration production
Vizualsite LLC
 
Getting Started with DDS in C++, Java and Scala
Angelo Corsaro
 
Materi SQL Query Berjenjang
Ngurah Endra
 

What's hot (10)

PPT
Banco de dados de grafos
Priscila Mayumi
 
PPTX
Domain Driven Design(DDD) Presentation
Oğuzhan Soykan
 
PDF
Fast DDS Features & Tools
eProsima
 
PDF
Oracle Database Appliance Workshop
MarketingArrowECS_CZ
 
PPTX
Siligong.Data - May 2021 - Transforming your analytics workflow with dbt
Jon Su
 
PPTX
OCI Overview
Kamil Wieczorek
 
PPTX
Domain driven design
Amit Mukherjee
 
PPTX
Oracle Office Hours - Exposing REST services with APEX and ORDS
Doug Gault
 
PDF
Snowflake free trial_lab_guide
slidedown1
 
PDF
Linked Data and Knowledge Graphs -- Constructing and Understanding Knowledge ...
Jeff Z. Pan
 
Banco de dados de grafos
Priscila Mayumi
 
Domain Driven Design(DDD) Presentation
Oğuzhan Soykan
 
Fast DDS Features & Tools
eProsima
 
Oracle Database Appliance Workshop
MarketingArrowECS_CZ
 
Siligong.Data - May 2021 - Transforming your analytics workflow with dbt
Jon Su
 
OCI Overview
Kamil Wieczorek
 
Domain driven design
Amit Mukherjee
 
Oracle Office Hours - Exposing REST services with APEX and ORDS
Doug Gault
 
Snowflake free trial_lab_guide
slidedown1
 
Linked Data and Knowledge Graphs -- Constructing and Understanding Knowledge ...
Jeff Z. Pan
 
Ad

Viewers also liked (20)

PDF
The Data Distribution Service Tutorial
Angelo Corsaro
 
PDF
Advanced OpenSplice Programming - Part II
Angelo Corsaro
 
PDF
Tuning and Troubleshooting OpenSplice DDS Applications
Angelo Corsaro
 
PDF
The DDS Tutorial Part II
Angelo Corsaro
 
PDF
Classical Distributed Algorithms with DDS
Angelo Corsaro
 
PDF
Desktop, Embedded and Mobile Apps with Vortex Café
Angelo Corsaro
 
PDF
Vortex Tutorial -- Part I
Angelo Corsaro
 
PDF
Vortex Tutorial Part II
Angelo Corsaro
 
PDF
Stream Processing with DDS and CEP
Angelo Corsaro
 
PDF
Building Real-Time Web Applications with Vortex-Web
Angelo Corsaro
 
PDF
Connected Mobile and Web Applications with Vortex
Angelo Corsaro
 
PDF
DDS Tutorial -- Part I
Angelo Corsaro
 
PDF
Getting Started with OpenSplice DDS Community Ed.
Angelo Corsaro
 
PDF
OpenSplice DDS v6
Angelo Corsaro
 
PDF
Getting Started with Vortex
Angelo Corsaro
 
PDF
Building Reactive Applications with DDS
Angelo Corsaro
 
PDF
Introducing Vortex Lite
Angelo Corsaro
 
PDF
The Data Distribution Service Tutorial
Angelo Corsaro
 
PDF
Building and Scaling Internet of Things Applications with Vortex Cloud
Angelo Corsaro
 
PDF
DDS QoS Unleashed
Angelo Corsaro
 
The Data Distribution Service Tutorial
Angelo Corsaro
 
Advanced OpenSplice Programming - Part II
Angelo Corsaro
 
Tuning and Troubleshooting OpenSplice DDS Applications
Angelo Corsaro
 
The DDS Tutorial Part II
Angelo Corsaro
 
Classical Distributed Algorithms with DDS
Angelo Corsaro
 
Desktop, Embedded and Mobile Apps with Vortex Café
Angelo Corsaro
 
Vortex Tutorial -- Part I
Angelo Corsaro
 
Vortex Tutorial Part II
Angelo Corsaro
 
Stream Processing with DDS and CEP
Angelo Corsaro
 
Building Real-Time Web Applications with Vortex-Web
Angelo Corsaro
 
Connected Mobile and Web Applications with Vortex
Angelo Corsaro
 
DDS Tutorial -- Part I
Angelo Corsaro
 
Getting Started with OpenSplice DDS Community Ed.
Angelo Corsaro
 
OpenSplice DDS v6
Angelo Corsaro
 
Getting Started with Vortex
Angelo Corsaro
 
Building Reactive Applications with DDS
Angelo Corsaro
 
Introducing Vortex Lite
Angelo Corsaro
 
The Data Distribution Service Tutorial
Angelo Corsaro
 
Building and Scaling Internet of Things Applications with Vortex Cloud
Angelo Corsaro
 
DDS QoS Unleashed
Angelo Corsaro
 
Ad

Similar to 20 Tips for OpenSplice Newbies (20)

PDF
Interoperable, Extensible and Efficient System Architectures
Angelo Corsaro
 
PDF
The OpenSplice DDS Revolution -- Episode II
Angelo Corsaro
 
PDF
DDS-PSM-Cxx and simd-cxx
Angelo Corsaro
 
PDF
Give or take a block
Tomer Perry
 
PDF
DDS Everywhere
Angelo Corsaro
 
PDF
File system
WajeehaBaig
 
DOCX
Best Answers.LESSON 10 1. What is the primary reason IPv6 h.docx
CruzIbarra161
 
DOC
systems_administrator test
Hunde Regasa
 
PDF
Parallels Plesk Panel 9 Quick Start Guide
webhostingguy
 
PDF
Security Research2.0 - FIT 2008
Raffael Marty
 
PPT
Os10
issbp
 
PPTX
Windows File Systems
primeteacher32
 
PDF
Amanda Enterpriseによるクラウドバックアップご紹介
OSSラボ株式会社
 
PPT
File Management
Munazza-Mah-Jabeen
 
PDF
Tarpm Clustering
day
 
PPTX
Administration & Gestion de Windows Server 2019
amouwawa
 
PPT
Ch12
NagellaRanjithKumar
 
PPT
file systems-CUSAT
Midhun Sankar
 
PDF
From Big GORM-centered into a cloud of fast redis nodes
Gailen Tecnologías
 
PDF
Registry
messyclick
 
Interoperable, Extensible and Efficient System Architectures
Angelo Corsaro
 
The OpenSplice DDS Revolution -- Episode II
Angelo Corsaro
 
DDS-PSM-Cxx and simd-cxx
Angelo Corsaro
 
Give or take a block
Tomer Perry
 
DDS Everywhere
Angelo Corsaro
 
File system
WajeehaBaig
 
Best Answers.LESSON 10 1. What is the primary reason IPv6 h.docx
CruzIbarra161
 
systems_administrator test
Hunde Regasa
 
Parallels Plesk Panel 9 Quick Start Guide
webhostingguy
 
Security Research2.0 - FIT 2008
Raffael Marty
 
Os10
issbp
 
Windows File Systems
primeteacher32
 
Amanda Enterpriseによるクラウドバックアップご紹介
OSSラボ株式会社
 
File Management
Munazza-Mah-Jabeen
 
Tarpm Clustering
day
 
Administration & Gestion de Windows Server 2019
amouwawa
 
Ch12
NagellaRanjithKumar
 
file systems-CUSAT
Midhun Sankar
 
From Big GORM-centered into a cloud of fast redis nodes
Gailen Tecnologías
 
Registry
messyclick
 

More from Angelo Corsaro (20)

PDF
Zenoh: The Genesis
Angelo Corsaro
 
PDF
zenoh: The Edge Data Fabric
Angelo Corsaro
 
PDF
Zenoh Tutorial
Angelo Corsaro
 
PDF
Data Decentralisation: Efficiency, Privacy and Fair Monetisation
Angelo Corsaro
 
PDF
zenoh: zero overhead pub/sub store/query compute
Angelo Corsaro
 
PDF
zenoh -- the ZEro Network OverHead protocol
Angelo Corsaro
 
PDF
zenoh -- the ZEro Network OverHead protocol
Angelo Corsaro
 
PDF
Breaking the Edge -- A Journey Through Cloud, Edge and Fog Computing
Angelo Corsaro
 
PDF
Eastern Sicily
Angelo Corsaro
 
PDF
fog05: The Fog Computing Infrastructure
Angelo Corsaro
 
PDF
Cyclone DDS: Sharing Data in the IoT Age
Angelo Corsaro
 
PDF
fog05: The Fog Computing Platform
Angelo Corsaro
 
PDF
Programming in Scala - Lecture Four
Angelo Corsaro
 
PDF
Programming in Scala - Lecture Three
Angelo Corsaro
 
PDF
Programming in Scala - Lecture Two
Angelo Corsaro
 
PDF
Programming in Scala - Lecture One
Angelo Corsaro
 
PDF
Data Sharing in Extremely Resource Constrained Envionrments
Angelo Corsaro
 
PDF
The DDS Security Standard
Angelo Corsaro
 
PDF
The Data Distribution Service
Angelo Corsaro
 
PDF
RUSTing -- Partially Ordered Rust Programming Ruminations
Angelo Corsaro
 
Zenoh: The Genesis
Angelo Corsaro
 
zenoh: The Edge Data Fabric
Angelo Corsaro
 
Zenoh Tutorial
Angelo Corsaro
 
Data Decentralisation: Efficiency, Privacy and Fair Monetisation
Angelo Corsaro
 
zenoh: zero overhead pub/sub store/query compute
Angelo Corsaro
 
zenoh -- the ZEro Network OverHead protocol
Angelo Corsaro
 
zenoh -- the ZEro Network OverHead protocol
Angelo Corsaro
 
Breaking the Edge -- A Journey Through Cloud, Edge and Fog Computing
Angelo Corsaro
 
Eastern Sicily
Angelo Corsaro
 
fog05: The Fog Computing Infrastructure
Angelo Corsaro
 
Cyclone DDS: Sharing Data in the IoT Age
Angelo Corsaro
 
fog05: The Fog Computing Platform
Angelo Corsaro
 
Programming in Scala - Lecture Four
Angelo Corsaro
 
Programming in Scala - Lecture Three
Angelo Corsaro
 
Programming in Scala - Lecture Two
Angelo Corsaro
 
Programming in Scala - Lecture One
Angelo Corsaro
 
Data Sharing in Extremely Resource Constrained Envionrments
Angelo Corsaro
 
The DDS Security Standard
Angelo Corsaro
 
The Data Distribution Service
Angelo Corsaro
 
RUSTing -- Partially Ordered Rust Programming Ruminations
Angelo Corsaro
 

Recently uploaded (20)

PPTX
VMware vSphere Foundation How to Sell Presentation-Ver1.4-2-14-2024.pptx
blackmambaettijean
 
PDF
Advanced methodologies resolving dimensionality complications for autism neur...
IAESIJAI
 
PDF
Electronic commerce courselecture one. Pdf
OmerMohamed64
 
PDF
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
PDF
Review of recent advances in non-invasive hemoglobin estimation
IAESIJAI
 
PPT
Teaching material agriculture food technology
LiaRayya
 
PPTX
Effective Security Operations Center (SOC) A Modern, Strategic, and Threat-In...
ReZa AdineH
 
PDF
Diabetes mellitus diagnosis method based random forest with bat algorithm
IAESIJAI
 
PPTX
ACSFv1EN-58255 AWS Academy Cloud Security Foundations.pptx
23bcla24
 
PDF
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Peter Tan
 
PDF
KodekX | Application Modernization Development
KodekX
 
PDF
TokAI - TikTok AI Agent : The First AI Application That Analyzes 10,000+ Vira...
SOFTTECHHUB
 
PDF
Empathic Computing: Creating Shared Understanding
Mark Billinghurst
 
PDF
Optimiser vos workloads AI/ML sur Amazon EC2 et AWS Graviton
Julien SIMON
 
PDF
Machine learning based COVID-19 study performance prediction
IAESIJAI
 
PDF
Network Security Unit 5.pdf for BCA BBA.
Serpent6
 
PDF
Approach and Philosophy of On baking technology
30anthuong17
 
PDF
NewMind AI Weekly Chronicles - August'25 Week I
NewMind AI
 
PDF
Chapter 3 Spatial Domain Image Processing.pdf
Getnet Tigabie Askale -(GM)
 
PDF
Unlocking AI with Model Context Protocol (MCP)
Brian McKeiver
 
VMware vSphere Foundation How to Sell Presentation-Ver1.4-2-14-2024.pptx
blackmambaettijean
 
Advanced methodologies resolving dimensionality complications for autism neur...
IAESIJAI
 
Electronic commerce courselecture one. Pdf
OmerMohamed64
 
Dropbox Q2 2025 Financial Results & Investor Presentation
Dropbox
 
Review of recent advances in non-invasive hemoglobin estimation
IAESIJAI
 
Teaching material agriculture food technology
LiaRayya
 
Effective Security Operations Center (SOC) A Modern, Strategic, and Threat-In...
ReZa AdineH
 
Diabetes mellitus diagnosis method based random forest with bat algorithm
IAESIJAI
 
ACSFv1EN-58255 AWS Academy Cloud Security Foundations.pptx
23bcla24
 
Architecting across the Boundaries of two Complex Domains - Healthcare & Tech...
Peter Tan
 
KodekX | Application Modernization Development
KodekX
 
TokAI - TikTok AI Agent : The First AI Application That Analyzes 10,000+ Vira...
SOFTTECHHUB
 
Empathic Computing: Creating Shared Understanding
Mark Billinghurst
 
Optimiser vos workloads AI/ML sur Amazon EC2 et AWS Graviton
Julien SIMON
 
Machine learning based COVID-19 study performance prediction
IAESIJAI
 
Network Security Unit 5.pdf for BCA BBA.
Serpent6
 
Approach and Philosophy of On baking technology
30anthuong17
 
NewMind AI Weekly Chronicles - August'25 Week I
NewMind AI
 
Chapter 3 Spatial Domain Image Processing.pdf
Getnet Tigabie Askale -(GM)
 
Unlocking AI with Model Context Protocol (MCP)
Brian McKeiver
 

20 Tips for OpenSplice Newbies

  • 1. 20 Tips for OpenSplice Newbies Angelo Corsaro, Ph.D. Chief Technology Officer OMG DDS Sig Co-Chair PrismTech angelo.corsaro@prismtech.com
  • 3. Tip #0 Domains, Partitions, Topics Domain (e.g. Domain 123) ¨ All DDS applications publish Copyright  2010,  PrismTech  –    All  Rights  Reserved. and subscribe data on a Partition belonging to a given domain Partitions (e.g. Partition “Telemetry”) ¨ Partitions are defined by means for strings and can Topic Instances/ be matched with regular Samples expressions ¨ If not explicitly specified the default partition in the default domain is automatically chosen
  • 4. Tip #0 Partitions Matching "building-1.floor-3.room-51" Copyright  2010,  PrismTech  –    All  Rights  Reserved. "building-1.floor-1.room-*" ... building-1.floor-3.room-5 ... ¨ Partitions are defined by means building-1.floor-1.room-111 building-1.floor15-.room-51 for strings and can be matched with building-1.floor-3.room-11 regular building-1.floor-1.room-1 building-1.floor-10.room-100 expressions] ... "building-1.floor-*.room-11?" Domain
  • 6. Tip #1 Choosing a Domain ¨ In OpenSplice DDS v5.x a domain is selected by: Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ Defining the OSPL_URI environment variable ¨ Passing a URI pointing at the the domain XML configuration file at OpenSplice startup ¨ Passing the URI of the configuration file as a string parameter of the DomainParticipantFactory::create_participant method ¨ Passing the name of the domain as specified in the configuration file to the DomainParticipantFactory::create_participant method ¨ Passing the empty string to represent the default domain to the DomainParticipantFactory::create_participant method
  • 7. Tip #1 Choosing a Domain Domain specified via URI Copyright  2010,  PrismTech  –    All  Rights  Reserved. Defining Domain Configuration at Startup dpf.create_participant( “file:///some/path/myospl.xml”, $ ospl start file:///some/path/myospl.xml qos, listener, mask); Default Domain dpf.create_participant( Domain specified via a Domain Name “”, qos, dpf.create_participant( listener, “MyDomainNameAsSpecifiedOnTheXMLFile”, mask); qos, listener, mask);
  • 8. Tip #1 Domains on OpenSplice v6.x ¨ The DDS specification did not originally define the type for the Copyright  2010,  PrismTech  –    All  Rights  Reserved. DomainId, as a result vendors where free to choose their on types ¨ As the DDSI/RTPS specification defines a DomainId with and integer it makes sense to uniform the DDS API to use an integer DomainID ¨ As a result, starting with OpenSplice DDS the domain will be selected specifying its associated id: dpf.create_participant( 15, qos, listener, mask);
  • 9. Tip #2 Start OpenSplice First Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ OpenSplice v5.x runs by default on a shared/ memory + daemon configuration ¨ As such, if you forget to start the infrastructure your application will fail at start up ¨ Thus always recall to run: $ ospl start
  • 10. Tip #3 Shared Memory Size OpenSplice DDS shared memory size is defined in its Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ configuration file. The size defined by the default configuration file is 10MBytes ¨ Beware that different OS have different limitations w.r.t. the maximum shared memory segment that can be allocated ¨ If you want to go beyond the OS limits you need to change the configuration of your kernel
  • 11. Tip #3 Linux ¨ The default value for the maximum shared memory segment is 32MBytes Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ This default can be changed in several ways (1) Adding this line to your /etc/rc.d/rc.local file: echo “your_max_shared_memory_size” > /proc/sys/kernel/shmmax (2) Changing the settings for the sys-limits (save changes on /etc/ sysctl.conf to maintain them across reboots): $ sysctl -w kernel.shmmax=yourMaxValue
  • 12. Tip #3 Windows The default maximum size for Shared Memory Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ segments on Windows is 2GB ¨ To exend it, say to 3GB, add the /3GB the boot.ini as shown below: [boot loader] timeout=30 default=multi(0)disk(0)rdisk(0)partition(1)WINDOWS [operating systems] multi(0)disk(0)rdisk(0)partition (1)WINDOWS="Windows NT Workstation Version 4.00" /3GB
  • 14. Tip #4 Topic Types & Keys Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ Topic Types can define some of their attributes as keys ¨ Yet, even when a Topic type does not define a key the keylist directive has to be provided -- just to tell the IDL compiler that this is a topic
  • 15. Tip #4 Topic Types & Keys Keyful Keyless Copyright  2010,  PrismTech  –    All  Rights  Reserved. enum TemperatureScale { CELSIUS, FAHRENHEIT, struct UniversalConstants { KELVIN double PI; }; double e; // Nepero Number double g; // Gravitational Constant struct TempSensorType { double NA; // Avogadro Number short id; double F; // Faraday Constant float temp; double K; // Boltzman Constant float hum; double c; // Light Speed TemperatureScale scale; }; }; #pragma keylist UniversalConstants #pragma keylist TempSensorType id
  • 16. Tip #5 Topic Instances Copyright  2010,  PrismTech  –    All  Rights  Reserved. enum TemperatureScale { CELSIUS, Writer Reader FAHRENHEIT, KELVIN }; 2 26.0 70.0 CELSIUS 1 26.0 70.0 CELSIUS struct TempSensorType { 1 26.0 70.0 CELSIUS 2 26.0 70.0 CELSIUS short id; float temp; float hum; TemperatureScale scale; }; #pragma keylist TempSensorType
  • 17. Tip #5 Topic Instances Copyright  2010,  PrismTech  –    All  Rights  Reserved. enum TemperatureScale { CELSIUS, Writer Reader FAHRENHEIT, KELVIN }; 2 26.0 70.0 CELSIUS struct TempSensorType { 1 26.0 70.0 CELSIUS 2 26.0 70.0 CELSIUS 1 26.0 70.0 CELSIUS short id; float temp; float hum; TemperatureScale scale; }; #pragma keylist TempSensorType id
  • 18. Tip #6 Topic Instances Lifecycle struct VehiclePosition { string plate; Copyright  2010,  PrismTech  –    All  Rights  Reserved. long x; long y; }; com.myco.VPos #pragma keylist VehiclePosition plate struct VehiclePosition { string plate; long x; long y; }; #pragma keylist VehiclePosition com.myco.VPos
  • 19. Tip #6 Topic Instances Lifecycle struct VehiclePosition { string plate; New Copyright  2010,  PrismTech  –    All  Rights  Reserved. long x; “A01” 100 200 long y; }; com.myco.VPos #pragma keylist VehiclePosition plate struct VehiclePosition { string plate; long x; long y; New }; #pragma keylist VehiclePosition “A01” 100 200 com.myco.VPos
  • 20. Tip #6 Topic Instances Lifecycle struct VehiclePosition { string plate; New Copyright  2010,  PrismTech  –    All  Rights  Reserved. long x; “A01” 100 200 long y; }; New #pragma keylist VehiclePosition plate “B41” 57 31 com.myco.VPos struct VehiclePosition { string plate; long x; long y; }; #pragma keylist VehiclePosition “B41” 57 31 “A01” 100 200 com.myco.VPos
  • 21. Tip #6 Topic Instances Lifecycle struct VehiclePosition { string plate; New Copyright  2010,  PrismTech  –    All  Rights  Reserved. long x; “A01” 110 210 “A01” 100 200 long y; }; New #pragma keylist VehiclePosition plate “B41” 57 31 com.myco.VPos struct VehiclePosition { string plate; long x; long y; }; #pragma keylist VehiclePosition “A01” 110 210 “B41” 57 31 “A01” 100 200 com.myco.VPos
  • 22. Tip #6 Topic Instances Lifecycle New Copyright  2010,  PrismTech  –    All  Rights  Reserved. “A01” 120 220 “A01” 110 210 “A01” 100 200 New “B41” 57 31 com.myco.VPos “A01” 120 220 “A01” 110 210 “B41” 57 31 “A01” 100 200 com.myco.VPos
  • 23. Tip #6 Topic Instances Lifecycle New Copyright  2010,  PrismTech  –    All  Rights  Reserved. “A01” 120 220 “A01” 110 210 “A01” 100 200 New “B41” 47 19 “B41” 57 31 com.myco.VPos “B41” 47 19 “A01” 120 220 “A01” 110 210 “B41” 57 31 “A01” 100 200 com.myco.VPos
  • 24. Tip #6 Topic Instances Lifecycle New Copyright  2010,  PrismTech  –    All  Rights  Reserved. “A01” 120 220 “A01” 110 210 “A01” 100 200 New “B41” Disposed - - “B41” 47 19 “B41” 57 31 com.myco.VPos “B41” 47 19 “A01” 120 220 “A01” 110 210 “B41” 57 31 “A01” 100 200 com.myco.VPos
  • 25. Tip #7 Default Lifecycle Settings The WriterDataLifecycle controls when instances are Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ disposed. By default DDS disposes unregistered instances ¨ Automatically disposing an instance perhaps is not what your want to do when terminating your application, as this would remove persistent data for the given instance!
  • 26. QoS
  • 27. Tip #8 Understand the QoS Model ¨ DDS defines 22 QoS DURABILITY LIVELINESS DEST. ORDER TIME-BASED FILTER policies that can be HISTORY OWENERSHIP PARTITION RESOURCE LIMITS Copyright  2010,  PrismTech  –    All  Rights  Reserved. LIFESPAN OWN. STRENGTH PRESENTATION applied to RELIABILITY DW LIFECYCLE communication USER DATA DEADLINE DR LIFECYCLE entities to control their TOPIC DATA GROUP DATA LATENCY BUDGET TRANSPORT PRIO ENTITY FACTORY local as well as end- to-end behaviour RxO QoS Local QoS ¨ Most of the QoS Policies that control an end-to-end property follow the so-called Request vs. Offered (RxO) Model based on which the QoS requested by the Consumer should not exceed the QoS Provided by the Producer
  • 28. Tip #9 History + Reliability Interplay ¨ The History QoS controls the number of samples that Copyright  2010,  PrismTech  –    All  Rights  Reserved. are maintained by DDS for a given topic ¨ DDS can keep the last n samples or keep all samples up to when they are not taken by the application ¨ The History setting has an impact on the reliability of data delivery as perceived by the application. Thus beware of your settings!
  • 29. Tip #9 History + Reliability Interplay struct Counter { QoS Settings int cID; Reliability = Reliable Copyright  2010,  PrismTech  –    All  Rights  Reserved. int count; }; #pragma keylist Counter cID History = KeepLast(1) History Depth = 1 History Depth = 1 (DDS Default) (DDS Default) Network 1 1 1 2 DataReader 2 1 1 2 2 3 DataWriter 3 1 2 2 2 3 3 1 Topic Topic DataReader Cache DataWriter Cache
  • 30. Tip #9 History + Reliability Interplay struct Counter { QoS Settings int cID; Reliability = Reliable Copyright  2010,  PrismTech  –    All  Rights  Reserved. int count; }; #pragma keylist Counter cID History = KeepLast(1) History Depth = 1 History Depth = 1 (DDS Default) (DDS Default) Network 1 2 1 2 DataReader 2 2 2 3 DataWriter 3 1 2 3 3 1 Topic Topic DataReader Cache DataWriter Cache
  • 31. Tip #9 History + Reliability Interplay struct Counter { QoS Settings int cID; Reliability = Reliable Copyright  2010,  PrismTech  –    All  Rights  Reserved. int count; }; #pragma keylist Counter cID History = KeepLast(1) History Depth = 1 History Depth = 1 (DDS Default) (DDS Default) Network 1 2 1 2 DataReader 2 3 2 3 DataWriter 3 1 3 1 Topic Topic DataReader Cache DataWriter Cache
  • 32. Tip #7 #10 Define Resource Limits ¨ DDS Provides a QoS that allow to control the amount of resources used by Copyright  2010,  PrismTech  –    All  Rights  Reserved. DataReaders and DataWriters ¨ By default, DDS does not imposes any limit, with the results that if you have a buggy application or an asymmetry in your system you might end-up consuming unbounded amount of memory -- and in the OpenSplice DDS case filling the Shared Memory ¨ To avoid this problem, always set appropriate Resource Limits for your application by defining: ¨ max_samples ¨ max_instances ¨ max_samples_per_instance
  • 34. Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; ¨ DataReader::read }; iterates over the available History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID sample instances ¨ Samples are not removed from the local cache as 1 1 1 2 1 3 1 4 result of a read DataReader 2 1 2 2 2 3 3 1 3 2 3 3 3 4 3 5 ¨ Read samples can be Topic read again, by accessing the cache with the Samples Read Samples not Read proper options (more DataReader Cache later)
  • 35. Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; ¨ DataReader::read }; iterates over the available History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID sample instances ¨ Samples are not removed from the local cache as 1 1 1 2 1 3 1 4 result of a read DataReader 2 1 2 2 2 3 3 1 3 2 3 3 3 4 3 5 ¨ Read samples can be Topic read again, by accessing Samples not Read the cache with the Samples Read proper options (more DataReader Cache later)
  • 36. Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; ¨ DataReader::read }; iterates over the available History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID sample instances ¨ Samples are not removed from the local cache as 1 1 1 2 1 3 1 4 result of a read DataReader 2 1 2 2 2 3 3 1 3 2 3 3 3 4 3 5 ¨ Read samples can be Topic read again, by accessing Samples Read Samples not Read the cache with the proper options (more DataReader Cache later)
  • 37. Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; }; ¨ DataReader::take History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID iterates over the available sample instances 1 1 1 2 1 3 1 4 Taken Samples are 2 1 2 2 2 3 ¨ DataReader 3 1 3 2 3 3 3 4 3 5 removed from the Topic local cache as result Samples not Taken of a take DataReader Cache
  • 38. Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; }; ¨ DataReader::take History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID iterates over the available sample instances 1 2 1 3 1 4 Taken Samples are 2 2 2 3 ¨ DataReader 3 2 3 3 3 4 3 5 removed from the Topic local cache as result Samples not Taken of a take DataReader Cache
  • 39. Tip #11 Read vs. Take struct Counter { int cID; QoS Settings int count; }; ¨ DataReader::take History = KeepLast(k) Copyright  2010,  PrismTech  –    All  Rights  Reserved. #pragma keylist Counter cID iterates over the available sample instances 1 3 1 4 Taken Samples are 2 3 ¨ DataReader 3 3 3 4 3 5 removed from the Topic local cache as result Samples not Taken of a take DataReader Cache
  • 40. Tip #12 Sample, Instance, View State History Depth = 2 ¨ Along with data samples, DataReaders provides state information allowing to detect relevant Copyright  2010,  PrismTech  –    All  Rights  Reserved. 1 1 transitions in the life-cycle of data as well as data writers 2 2 3 ¨ Sample State (READ | NOT_READ): Determines DataReader SampleInfo wether a sample has already been read by this DataReader or not. 1 1 1 2 ¨ Instance State (ALIVE, NOT_ALIVE, DISPOSED). 2 2 2 3 Determines wether (1) writer exist for the specific 3 1 instance, or (2) no matched writers are currently Samples available, or (3) the instance has been disposed Topic ¨ View State (NEW, NOT_NEW). Determines wether this is the first sample of a new (or re-born) instance DataReader Cache
  • 41. Tip #13 Beware of Invalid Samples! ¨ For each data sample accessed via a read or take Copyright  2010,  PrismTech  –    All  Rights  Reserved. DDS provides you with a SampleInfo ¨ The SampleInfo contains meta-information about the Sample, such as timestamp, lifecycle information, etc., but most importantly tells you if the data is valid or not! ¨ Data is not valid, when the sample you are receiving notifies things like an instance being unregistered or disposed
  • 42. Tip #14 Reading Only “Fresh” Data Copyright  2010,  PrismTech  –    All  Rights  Reserved. dr.read(samples, infos, LENGTH_UNLIMITED, // read all available samples NOT_READ_SAMPLE_STATE, ANY_VIEW_STATE, ALIVE_INSTANCE_STATE);
  • 43. Tip #15 Reading All Data Copyright  2010,  PrismTech  –    All  Rights  Reserved. dr.read(samples, infos, LENGTH_UNLIMITED, // read all available samples ANY_SAMPLE_STATE, ANY_VIEW_STATE, ALIVE_INSTANCE_STATE);
  • 44. Tip #16 Getting Everything Copyright  2010,  PrismTech  –    All  Rights  Reserved. dr.read(samples, infos, LENGTH_UNLIMITED, // read all available samples ANY_SAMPLE_STATE, ANY_VIEW_STATE, ANY_INSTANCE_STATE); NOTE: As explained on the Tip #13 in this case you might get invalid data samples, thus have to check on their validity via the SampleInfo.valid_data attribute
  • 45. Tip #17 Status vs. Read Condition Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ Both a Status as well as a Read Condition can be used to wait for data to be available on a DataReader ¨ The main difference is that a ReadCondition allows to set the exact SAMPLE, VIEW and INSTANCE status for which the condition should trigger, while the StatusCondition triggers when a sample is received
  • 46. Tip #17 Status vs. Read Condition Guidelines Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ If what you really care is getting a condition that triggers for the state ANY_SAMPLE_STATE, ANY_VIEW_STATE and ANY_INSTANCE_STATE than use a StatusCondition as this is more efficient than a ReadCondition ¨ If you are interested in having a condition that triggers for specific a status then use the ReadCondition
  • 48. Tip #18 Return Memory Loans Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ The DataReader read/take allocate loan memory to the application when the length of the containers passed for storing samples and info is zero ¨ In this case the loaned memory must be returned via a return_loan operation!
  • 49. Return Memory Loans Tip #18 Copyright  2010,  PrismTech  –    All  Rights  Reserved.
  • 50. Tip #19 Beware of Strings Ownership ¨ The DDS C++ API takes ownership of the string you pass Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ As a result, you need to understand when it is necessary to “duplicate” a string ¨ To this end, DDS provides the DDS:string_dup call to facilitate this task subQos.partition.name.length (1); subQos.partition.name[0] = DDS::string_dup (read_partition);
  • 52. Tip #20 Read, Write, Ask Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¨ Read the Manuals and if possible the Specification ¨ Write your own code examples ¨ Don’t be shy to ask questions on the OpenSplice mailing list
  • 53. :: Connect with Us :: Copyright  2010,  PrismTech  –    All  Rights  Reserved. ¥ opensplice.com ¥ forums.opensplice.org ¥ @acorsaro ¥ opensplice.org ¥ opensplicedds@prismtech.com ¥ @prismtech ¥ crc@prismtech.com ¥ sales@prismtech.com ¥ youtube.com/opensplicetube ¥ slideshare.net/angelo.corsaro