Azure based software architecture

Hi,
I would like to create a web based application. It would be a kind of statistics. The user can upload files into the software, the data is processed and later the UI displays the processed data.
I realized that I need a web role for the website. A worker role to process the files. A database to store the processed data.
What I do not really know which communication channel to use. I've checked the service bus but the message is limited to 64 kb (48). There is a blob storage as well for larger files.
Should the web instance save the files into a blob and then notify the worker role via service bus that there is a new file to be processed?
Sorry for the beginner question but I would like to make sure that I start with the best architecture.
Could you please review my architecture and let me know whether it should be changed or not?
Thanks in advance.

uploading to blob storage and writing to a queue to notify a worker role or even an Azure WebJobs should be fine.
if you would like to look at common cloud design patterns, i suggest you have a look at :
Cloud Design Patterns

Similar Messages

Software architecture to utilize XML DB

Hi,
We're using currently using Toplink 9.0.4 with an Oracle 9i DB.
I'm interested to find out what level of support there is in Toplink for the "new" XML DB features (in the Oracle 10g database for example).
The material I've seen so far on the Toplink developer preview suggests that Toplink will manage mappings from an object model to a XML representation.....but that XML representation is purely an "in-memory" - i.e. it does not generate the "SQL" to store objects in an XML DB.
There seems to be a gap in the way XML data is managed and the way Java domain models (and persistence ) are typically architected.
The way we craft our services at the moment is based upon Java objects which encapsulate a request. Changing the contents of a request (e.g. adding a new attribute), breaks existing implementations that utilize that service (because of Java's serialization model).
I'd like to define a service which takes a lump of metadata (passed as a generic object model, e.g. XML), but also allows that metadata to be evolving/variable/dynamic . The initial service implementation would look up a Java business object based upon the request information...include any transaction "attributes" from the request and persist it. Nothing new there, but what if I want to allow clients to attach "custom attributes" to their requests. The traditional model of request --> Java Business Object --> relational database seems to break down. Java Business Objects aren't typically very good at handling dynamic addition of properties.
Does there exist a software architecture that allows for Java Business Object "properties" to be dynamically attached and persisted.
It seems that the big push on XML is heading towards that goal....but at the moment I can't see how the pieces would actually fit together ......
Apologies that my ramblings are a bit wild!
Regards
Mike

TopLink 10.1.3 added XML support in three ways.
- Object-XML mapping - Support was added to be able to map Java object to XML documents. This allows for in-memory two-way conversion between object and XML, and usage with XML messaging and web services.
- Oracle XDB XML Type - Support was added to store a plain XML attribute of a Java object mapped in relational TopLink to an Oracle XDB XML type field.
- EIS - Support was added to allow mapping Java objects to XML records and Interactions in a non-relational EIS system through a JCA adapter that supports XML records.
From the sounds of things you may wish to:
Use TopLink OX to map your XML messages to your Java object model. For the dynamic XML content you could use a Transformation mapping in TopLink OX to store this dynamic content in a plain XML field in the object. You could then use Oracle XDB and the TopLink XMLType mapping to store this dynamic XML content in an XMLType field in the database.

Software architecture for multiple device interface

Hi everybody!
I'm quite new at LabVIEW and I have a question regarding software architecture.
I have to program a software for controlling 3 motorized stage, a camera and a pump. I have separately programmed VI's that control each of them and now I would like to merge them in a single program and I don't know what is the best way to do it.
So far my sub-VIs are made of "event-based" structure that reacts when the users click on pushbuttons. Should I simply put all my sub-VIs (i.e. multiple event structures) in one big while loop? Is it the best way to do it? I've been looking for a while for references about that but I couldn't find any relevant ones and I would really appreciate your help. Also if you know good references please share them to me.
Many thanks!
Bests,
Julien

If you could give me more details or send me a link to references about it it would be awesome.
Sure, have a look at the attachement and see if it could be extended for your use?
It is a fairly simple implementation. I use it for reading for a number of different InputStream devices, such as FPGA DMA FIFO's, sockets, files, etc..
Br,
/Roger
Attachments:
LV2009 InputStreams.zip ‏1694 KB

HOW DO I OPEN WINDOWS XP ON MY NEW IMAC TO USE A WINDOWS BASED SOFTWARE?

HOW DO I OPEN WINDOWS XP ON MY NEW IMAC TO USE A WINDOWS BASED SOFTWARE?

Assuming you already have it installed: Go to system preferences and chose 'start up disc' in the system menu. Then select the Windoze partition and restart it. - To get back to OSX, find the Boot Camp icon in the tool bar bottom right.

Re: Container Based Software Configuration Management

I've read the overview of Container Based Software Configuration Management and
have to admit that it is a nice concept. In fact, our SCM product SiberSafe Expert
2.0 (www.siberlogic.www) almost literally follows the guidelines mentioned in
the overview except that we call it Component Management.
SiberSafe recognizes the fact that most of software projects consist of software
components that have been inherited from and are being reused in other projects.
SiberSafe was built on the premise that it is critical to be able to version components
consisting of multiple files and folders so that a new project can be created
out of particular versions of multiple exising
components.
SiberSafe also recognizes the fact that various versions of various components
may be distributed not only across multiple projects, but also across multiple
lines of development (we call them "development streams"). This is why SiberSafe
makes the propagation of new versions of components from their "home" streams
into other streams where they are used and reused
as simple as a mouse button click.
When a few projects have a few development lines (streams) each and there are
tens or even hundreds of software components reused across these projects and
streams, it becomes pretty difficult to keep track of what went where and when.
What is the version of a specific java bean that our project currently uses? Is
there a newer version for that bean available? What verson of that bean do other
projects use? In order to provide users with
clear answers to this kind of questions SiberSafe automatically generates a visual
development flowchart that graphically shows relationships and interactions among
various components, component versions, projects and lines of development.
In fact, SiberSafe can easily generate such a flowchart even for a single file,
showing the full "travel" history of that file across projects and streams.
Alex Povzner,
SIBERLOGIC INC
www.siberlogic.com
905-4742384
Hans Thelosen <[email protected]> wrote in message
news:[email protected]...
On 16 and 17 May in Soesterberg (the Netherlands) a tool-independent
software configuration management training will be held.
In the last couple of years Software Configuration Management has grown toa
mature Software Engineering discipline with its own Functions, Methods and
Techniques. Container Based Software Configuration Management is a new
method for SCM that centralizes facilities for reuse and parallel
development of Software.
For more information see:
http://www.qa-systems.com/training/improvement_scm.jsp

All,
There has been some uptake and some great feedback on rConfig since I made this announcement two weeks ago. rConfig is offically released as Version1 today. Please login to www.rconfig.com and download a copy.
You know, it takes less than 1 hour to install rConfig on Linux and more than 2 hours to get some of it's well known commercial counterparts up and running. There is even a complete, easy-to-follow Linux build document designed especially for the rConfig installation on www.rconfig.com. You'll be backing up Cisco configuration and show outputs in no time with rConfig. And even learning a bit of Linux along the way.
And remember, it's free & open source.
Regards
Stephen
==========================
http://www.rConfig.com
A free, open source network device configuration management tool, customizable to your needs!

Correct software architecture

Hello.
I´m designing a complex application and I don´t know what software architecture is the best. I have been reading the "LabVIEW Intermediate I: Succesful Development Practices" manual. Lesson 3 gives the reader a lot of options for software architecture selection, but, although it gives a lot of explanations, I can´t find the correct one (maybe I should use a combination of some of them).
The application runs in a PC. This application communicates with 4 controllers via CAN and with another one via Ethernet.
Both communications are very asynchronous because I have to read a lot of data but send only a few.
I have to receive a lot of data from the first controllers (CAN connection) and display a graph containing all data. Reception must be as fast as possible without losing any data. During graphic displaying time is important but not critical. In the other hand, is critical not to lose any data. I mean it is critical to collect all data although they aren´t displayed in "Real time".
I have to send some data to the controllers via CAN but in this case, transmision rate can be quite slow, so I guess it can run in another loop.
Communication with the other controller (via Ethernet) have a similar behaviour, but it don´t have to be as fast as CAN communication.
Finally, I have a User Interface where basically I change parameters and send them to controllers so there are no intensive processing or activity.
I have been thinking about using Master/Slave or Producer/Consumer architectures but I don´t know what architecture best fits my application.
Can you help me?
Thanks!

Personally, I would stick with the producer/consumer approach for this in order to seperate out the UI code from the rest.
Additionally, I would launch a background process (out-of-line via VI server, but I can't find any decent references at the moment) for each CAN so that the communication for each device can run in a seperate thread. This way if one blocks up or generates an error, it won't block the others. These VIs running in the background then send the acquired data via QUEUE to the main program.
I've used this approach in the past for communicating in parallel with many devices and have found it to be very good. The time required to get it up and running is maybe a bit longer, but in the end it's worth it.
So basically, 1 VI 1 device on the level closest to the hardware and then one UI VI which communicates with the others via QUEUEs.
That's my 2c, maybe someone else has a different approach.
Additionally, Sorry I can't give an example of what I mean, or a decent reference but y quick saerch of the forum didn't show me anything useful. With time it'll work, but I don't have too much time at the moment.
Hope this helps
Shane.
Using LV 6.1 and 8.2.1 on W2k (SP4) and WXP (SP2)

Java software architecture - discussion forum

Hello,
I am looking for a forum where you can discuss software architecture topics. There are so many questions and controversal discussions in this field. I did not really find a forum for that in particular.
A few days ago I posted a question in the RMI-forum (by the way: with no answers so far) where I wanted to know how to distribute data from one source to many receivers (on the same JVM and also on different JVMs). The receiver should be able to select the data they need (so it not always the same data which is send to the different listeners.
My idea/suggestion was, to create this interface for example
public interface DataReceiver extends Remote {
public boolean isRemoteObject() throws RemoteException�
public void transmitData(MyData data);
}With that interface I am able to use RMI to access Objects on other JVMs as well as Object on the same JVM which are only implementing this interface.
Now I implement an ArrayList with DataReceiver-Objects and if data changes I can call the transmitData(..)-method of all these objects.
Is this a good way to do this, or are there better solutions. I am open to any help here.
I think this is not only a RMI-topic, because maybe a JMS-solution would be maybe a better solution. And why use RMI at all? Maybe a network/socket solution is faster.
By the way: Does anyone have experience in building a trading system (for algorithmic trading) ?
thanks
tk

Hello duffymo...
duffymo wrote:
1) Speed
If I use the RMI/interface-solution you can directly use the methods of the interface without having some layer in between if you are in the same JVM. This should be faster as JVM. I have no idea what the situation is like if the receiver is not on the same JVM. How fast is the RMI/interface-solution? How fast is the JMS-solution?RMI? Same JVM? No you're not - each remote client runs in its own JVM. What are you talking about? If they'er all in the same JVM, there's no need for RMI. That's just JavaBeans and events.o.k....from the start....some clients are on the same machine AND some are on different ones. Speed is essential here, especially for the clients on the same machine (components of the server itself). In using this interface you could manage clients no matter whether they are local objects or remote objects. The object that have to notify the clients even do not have to know whether the clients are on the same JVM or the data is forwarded via RMI (or sockets).
If the objects are local the "notification" is just a method call (therefore very fast) and if they are remote you have a pure RMI remote object call. Which also should be faster than JMS.
JMS, by definition, is asynchronous. The publisher sends off the message and doesn't wait for any response. It also doesn't worry about who responds, because it's up to subscribers to indicate their wish to receive the message.yes, so there is a lack of control, what I said. There many tasks where you need this, but there also many tasks where you cannot use MOM.
I have no idea why you're so worried about speed in that case....because this is the requirement !!!! I mentioned it several times. We are talking here about milliseconds. Each detour of the data flow or every data-copying (within the memory or even worse to a database/HDD) takes valuable time.
2) Control
the RMI/interface-solution provides more control. How can you say "more control" when you admit that you don't know JMS?You can get the idea of MOM (i.e. JMS) very fast I think. And I dealt with the topic for some time. I am working in the computer science for 17 years now, believe me, I am able to learn.
I�ve explored the acknowledge-methods of JMS and they all are very slow. You can not use this in a system where speed matters. And you get no direct feedback. (If receiver A and B subscribed to some data, how to find out that both got the data? The sender do not know its receivers and this is one source of the lack of control)
And: If you send the data via a method call, you can get a return value and exceptions. You are free to do (almost) eveything you want. Do you really want to say, that with JMS you have this kind of control?
The sender knows all the clients and can check, whether they have really received the message or not. Maybe even provide some feedback. Can you provide this with JMS?The whole idea behind publish/subscribe is that the publisher doesn't have to know.
JMS can be set up to guarantee delivery.Yes....BUT: I want to find the best solution to a GIVEN problem.
I think here speed is really what matters. And that is the main reason against using JMS. Therefore I think I do not have to deal with JMS any more. The idea of JMS is nice, and I will use surely use it in the future for some other projects, but here - I think - I cannot use it.
and something to "guarantee delivery" ... this is not always a good thing. In some cases the receiver do not want to receive "old data". It depends on the task to accomplish.
3) Messages
in the RMI-Interface-solution I can send almost all kind of objects to the client (serializable). As I can see you can also send java-objects via JMS, but it seems to be more complicated (messages have to match a standard (header, properties, body)...).Read more about JMS. Ditto for it if you send an ObjectMessage.???
And I said.....?????
I think the RMI/interface-solution is not easy. For example you have to use threads (incl. thread-pool) to send all the messages almost at the same time. On the other hand I think it is not too difficult and maybe worth the effort.RMI isn't easy, and it's brittle. Change the interface and all the clients change.yes, I agree. I faced a lot of problems with RMI. But then provide some alternatives. JMS seems to be too slow (and I do not have the time for coding an interface on the socket basis).
I think it is difficult and not worth the effort. JMO, of course.
JMS is done for you. You don't have to maintain that code.except the performance problem
The most important points are here reliability and speed. JMS is more reliable than anything you'll ever write, and it's fast enough."fast enough" .... how can you decide. I can tell you, it is not fast enough.
and reliable: yes I would prefer JMS here. really! I have to admit while reading the specs I began to like JMS. really. There are some chat-functions to implement. This should be very ease with JMS.
I think going down an RMI road for this would be foolish. I wouldn't do it.Maybe there are some other solutions. I will keep looking. And I will check the frameworks stefan.schulz mentioned.
I am very sorry, but I think this is a justified question. And this comment is not very nice at all. There are so many different areas in the Java-World, and no one can be expert in every field. I think this is what all the forums are about. Share knowledge and help others. It's not intended to be mean or demeaning. I don't know if you're aware that excellent commercial trading systems exist, written by people that wouldn't have to ask about RMI or JMS to do it. The fact that you have to ask suggests to me that you aren't one of those experts. If you're doing this as a learning exercise, I'd say go for it. Do it both ways in that case and learn for yourself. Just be realistic about the end goal: it won't be a system that will be used in production anywhere.I am in the financial business for almost 8 years now. And there are many different areas and roles. You have no idea what the systems are like, that I usually design (and they are very, very successful). I think your comments are rather unrefined. And maybe you should think about beeing a little bit more open-minded. You cannot prejudge everything.
Software development is very complex because of all the tools and methods in this field. Knowing when to use a tool/framework/method... is as important as knowing when NOT to use a tool/framework/method/...Yes, and knowing your limitations is also very important.what is the purpose of this last sentence?!
I really want to know!
You are the one, who wrote
JMS is more reliable than anything you'll ever write, and it's fast enough.Without knowing anything about the system (that is not true: I mentioned several times that speed is very important).
Without knowing anything about me...
I really appreciate your contribution here (it really helps), but I stick to it: it is not very nice to goof on someone.
regards
tk

Just got a macbook pro, no disc drive. how do I install disc-based software? Do I have to buy an external drive? ARgh.

Just got a macbook pro, no disc drive. how do I install disc-based software? Do I have to buy an external drive? ARgh.

Yes, an external drive will be necessary. You may purchase the Mac superdrive from the online store or one from another manufacturer. All will work.
http://store.apple.com/us/product/MD564ZM/A/apple-usb-superdrive
Ciao.

How to install DVD based software on Mini

How do I install DVD based software on a Mac Mini? Does it have a "CD/DVD" Drive? There appears to be a slot on the back. Is that a "SuperDrive" or whatever the **** they call it now?

If you have another PC/Mac on your network you can activate disc sharing to allow the Mac Mini to hijack the disc drive of that machine or acquire the Apple external drive or similar third party unit.
The current models have NO built in disc drive, the space is now reserved for the optional second SSD on select added value models, and can also be used to install your own second drive with little jiggery pokery.
With the likes of the Apple App Store, Amazon Store Microsoft Market (Office 2011) and Steam for games and other online vendors the need for physical discs is rapidly diminishing provided you have access to decent broadband bandwidth.
If you are a serious gamer then the mac mini is probably the wrong machine for you anyway (PC or console surely).
If you are using professional Adobe/Maya type software then the $70.00 external isn't going to break the bank is it ?

Any free coldusion based software deployment tools?

Hi, i'm trying to find a web based software deployment tool, ideally written in coldfusion. Does anyone know of any?
Thanks
Z

It's very easy with Time Machine (which is built into the OS). Just turn on Time Machine in System Preferences, plug in an external hard drive, and you should be asked whether you want to use that drive for TM. (Warning: this will erase the drive, so don't do this if there's data on the drive!)
For more about backups, see my Mac Backup Guide, as well as Pondini's excellent Time Machine FAQ.
(Note that my pages contain links to other pages that promote my services, and this should not be taken as an endorsement of my services by Apple.)

WAE 674 Disk based software not installed.

Hi,
Have an WAE 674 with 4.4.3c installed, with problem.
Using WCCP and I am able to logon to the device.
Have this msg and in in the CM it is down.
System Initialization Finished.
Disk based software not installed.
Some applications may not run correctly!
Any sugestion?
I have reload still problem.
Regards Jan Rockstedt

What steps did you take to resolve the issue? I have the same issue on WAVE 594 running 4.4.5.
I get the error message when I telnet into device.

I work for a kitchen design firm and want to know if the is a mac based software for kitchen and bath designs

I am looking for a Mac based software program that is easy to use to design kitchens and baths.
Do you know of any?

A friend uses Google's SketchUp ... free, and it created isometric views that satisfied the city/county/neighborhood-association when building his deck.
"Free" is worth a look, I would guess.

Cloud-based software-as-a-service (SaaS)

Hello
Does anyone know where can I find documentation on how to configure SaaS with WAAS?
I'm looking to setup salesforce application on the WAAS for WAN optimization out to the branch office. All I can find is whitepapers that talks about what it does. I need detail steps on how to properly configure it or examples showing how to.
Do you just build an SSL service connection to *.salesforce.com?
How will the traffic from the branch office be able to route to salesforce.com when they have their own local ISP?
Do you add static routes for all of the possible IP addresses for the application out of your Data Center? If so, this makes it very difficult to manage since the cloud service can change their IP address at any given time.
I would appreciate for someone to give me a simple example to follow. :-)
Thanks in advance!
John

Hello John,
I am not an expert on this but still PLEASE read the following on detailed, you will find most of your questions answered:
Simplified Deployment Model for Cloud-Based Services
Cloud-based software-as-a-service (SaaS) providers such as WebEx.com and Salesforce.com primarily use HTTPS to securely deliver services to their clients. Using Cisco WAAS SSL Application Optimizer, Cisco WAAS can optimize delivery of these services to the remote branch-office users who connect to these services through a backhaul connection to the data center.
However, the solution poses some unique challenges related to simplifying the implementation and deployment model. Typically, SaaS providers have multiple SSL server farms with multiple hosts spanning several data centers. When a client initiates an SSL connection request to a server located in the SaaS server farm, the Cisco WAAS SSL Application Optimizer needs to map the destination IP address in the incoming SSL request to an SSL accelerated service to present the right SSL certificate to the client to perform an SSL handshake.
For SSL services that are hosted in the enterprise data center, the IT administrator knows and controls the SSL server IP address and can provide it to the data center Cisco WAAS. But for an SSL service hosted at a third-party SaaS provider cloud, the SSL server IP address is not controlled by the IT administrator. Also, there may be not just one but multiple server IP addresses even for a single SaaS service, and these may be subject to change.
To simplify the deployment model for SaaS optimization, Cisco WAAS provides support for domain names in the SSL accelerated service configuration. This model can be applied to any cloud-based SaaS applications such as Salesforce.com and WebEx.com. For example, when an SSL accelerated service is configured in the data center Cisco WAE with a wildcard domain name (for instance, *.webex.com) option, the Cisco WAAS SSL Application Optimizer performs a reverse Domain Name System (DNS) lookup on the destination server IP upon receiving an SSL connection request from the client, and if the IP address resolves to a host that matches the configured wildcard domain name (for instance, *.webex.com), then the appropriate SSL accelerated service policy is applied to this connection (Figure 3).
Cisco Wide Area Application Services SSL Application Optimizer Deployment Guide (Simplified Deployment for SaaS Optimization)
http://www.cisco.com/en/US/prod/collateral/contnetw/ps5680/ps6870/deployment_guide_c07-541981.html
This topic is pretty complex for a support forum, I suggest to open a TAC case or contact your Cisco partner or PDI support.
hope this helps!
Felix

Tomcat software architecture

do you know where i can find tomcat's software architecture? architecture that is a drawing not words? If possibel do you also know where i can find the ejb or rmi software architecture? Dont know where the correct forum to place this so sorry if it is wrong thanks!

One option is to download the tomcat source and use one of these
tools to generate your own diagram:
http://www.cse.ucsc.edu/classes/cmps115/Fall02/tools/tools.html
This diagram...
http://www.onjava.com/onjava/2001/12/12/graphics/tc1.gif
...from this article...
http://www.onjava.com/pub/a/onjava/2001/12/12/openjms.html
...is probably too simplistic for your needs.
If class loader information is what you wanted to understand better,
there is a fantastic :-) ASCII diagram here:
http://jakarta.apache.org/tomcat/tomcat-4.0-doc/class-loader-howto.html
Other than that, it appears to me your best bet is the Tomcat docs...
http://jakarta.apache.org/tomcat/tomcat-5.0-doc/introduction.html
...including the API:
http://jakarta.apache.org/tomcat/tomcat-5.0-doc/catalina/docs/api/index.html
No architectural diagrams, but you may want to make a request
to the Tomcat documentation folks (refer to the bottom of the
the "introduction" link above). Be sure to give more detail
about what kind of diagram(s) you think would be most helpful
(since your post was a little light on those details).
Good luck!

ASA Software architecture Doc.

where can i find ASA documents related software architecture ?(inclued pictures)-_-
I have serched almost all of cisco.com , but couldn't find......

Son,
See the below links for what you are after:-
http://www.cisco.com/en/US/customer/products/ps6120/products_installation_and_configuration_guides_list.html
http://www.cisco.com/en/US/customer/products/ps6120/prod_configuration_examples_list.html
HTH>

Azure based software architecture

Similar Messages

Maybe you are looking for