[flexcoders] Flex Deployment in Production Environment
Hi all, Our 100% Flex application is deployed on to Apache Tomcat 5.0 with a Java (5.0) middle tier talking to a Mysql (5.0) backend. Everything is installed on a SuSe Linux box (currently SuSe Linux 10.0 but for production we are thinking of moving to Enterprise Server 9.0). We are currently trying to work out what is a reasonably cheap setup that can cater for 1000 concurrent users (and maybe 10 times that after a few months). Regarding Flex and Tomcat deployment, how can I start working out the RAM requirements for this? What if I add in another box with load balancing and failover capabilities? Any help would be much appreciated. Kam. -- Flexcoders Mailing List FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com Yahoo! Groups Links * To visit your group on the web, go to: http://groups.yahoo.com/group/flexcoders/ * To unsubscribe from this group, send an email to: [EMAIL PROTECTED] * Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
[flexcoders] Re: Auto detection of Flash Player Version
Hi Andrew,
Detection of old flash player version is fine.
What I am experiencing is this. I already have 8.0.22 player
installed. When the flex-config.xml file is detecting 7.0.14, the
application is fine. When i change it to 8.0.22 and redeploy, the
application doesn't work anymore. (The loading bar reaches the end,
and stops).
I change the flex-config.xml back to the original and everything works
again!
The only change is the flex-config.xml file. Nothing else!
Anybody with ideas on why this is happening?
Cheers,
Kam.
--- In flexcoders@yahoogroups.com, Andrew Spaulding [EMAIL PROTECTED] wrote:
Hi Kam,
It should be as simple as changing the following information in the
flex-config.xml
required-version8/required-version
!-- 'dot' release of the required player --
required-major-revision0/required-major-revision
!-- minor revision number of the required player --
required-minor-revision22/required-minor-revision
I had 7.0.35 installed and after changing the required version to
8.0.22 both Firefox and IE prompted me to upgrade my Flash Player.
Cheers,
Andrew Spaulding
www.flexdaddy.info
--- In flexcoders@yahoogroups.com, Kam-Wing Pang [EMAIL PROTECTED] wrote:
Hi,
Our application currently checks for the minimum flash player
version of
7.0.14 which is set in flex-config.xml:
flash-player
!-- enable/disable player detection --
enabletrue/enable
!-- require 7.0.14 --
!-- required player version --
required-version7/required-version
!-- 'dot' release of the required player --
required-major-revision0/required-major-revision
!-- minor revision number of the required player --
required-minor-revision14/required-minor-revision
!-- auto-install the player on windows and ie-only --
!-- When windows-auto-install is true, IE will attempt to
download
the ActiveX player without
user intervention. When false, the user is directed
to the
download url for upgrade. --
windows-auto-installtrue/windows-auto-install
!-- player upgrade page - users without the required Flash
version
will be directed here --
download-url{context.root}/flex-internal/detection-kit/upgrade_flash/upgra
de_flash.html/download-url
download-https-url{context.root}/flex-internal/secure/detection-kit/upgrad
e_flash/upgrade_flash.html/download-https-url
!-- activex upgrade page - the location for downloading the
ActiveX
player--
activex-download-urlhttp://download.macromedia.com/pub/shockwave/cabs/flas
h/swflash.cab/activex-download-url
activex-download-https-urlhttps://download.macromedia.com/pub/shockwave/ca
bs/flash/swflash.cab/activex-download-https-url
!-- plugin upgrade page - the location for downloading the
Plugin
version of the player --
plugin-download-urlhttp://www.macromedia.com/go/getflashplayer/plugin-dow
nload-url
plugin-download-https-urlhttps://www.macromedia.com/go/getflashplayer/plu
gin-download-https-url
/flash-player
I then changed the version number to be 8.0.22, redeploy my
application and
then my application doesn't load properly. The application shows the
initialisation bar, reaches the end and stops. When I revert back the
changes to check for 7.0.14, it works fine!
What is happening? How can a check for player version stop my
application
from loading? I have installed 8.0.22 which I double checked on the
macromedia website! This happens on both Firefox and IE browsers.
Any help is appreciated,
Kam.
--
Flexcoders Mailing List
FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com
Yahoo! Groups Links
* To visit your group on the web, go to:
http://groups.yahoo.com/group/flexcoders/
* To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
* Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
[flexcoders] Auto detection of Flash Player Version
Hi,
Our application currently checks for the minimum flash player version of
7.0.14 which is set in flex-config.xml:
flash-player
!-- enable/disable player detection --
enabletrue/enable
!-- require 7.0.14 --
!-- required player version --
required-version7/required-version
!-- 'dot' release of the required player --
required-major-revision0/required-major-revision
!-- minor revision number of the required player --
required-minor-revision14/required-minor-revision
!-- auto-install the player on windows and ie-only --
!-- When windows-auto-install is true, IE will attempt to download
the ActiveX player without
user intervention. When false, the user is directed to the
download url for upgrade. --
windows-auto-installtrue/windows-auto-install
!-- player upgrade page - users without the required Flash version
will be directed here --
download-url{context.root}/flex-internal/detection-kit/upgrade_flash/upgra
de_flash.html/download-url
download-https-url{context.root}/flex-internal/secure/detection-kit/upgrad
e_flash/upgrade_flash.html/download-https-url
!-- activex upgrade page - the location for downloading the ActiveX
player--
activex-download-urlhttp://download.macromedia.com/pub/shockwave/cabs/flas
h/swflash.cab/activex-download-url
activex-download-https-urlhttps://download.macromedia.com/pub/shockwave/ca
bs/flash/swflash.cab/activex-download-https-url
!-- plugin upgrade page - the location for downloading the Plugin
version of the player --
plugin-download-urlhttp://www.macromedia.com/go/getflashplayer/plugin-dow
nload-url
plugin-download-https-urlhttps://www.macromedia.com/go/getflashplayer/plu
gin-download-https-url
/flash-player
I then changed the version number to be 8.0.22, redeploy my application and
then my application doesn't load properly. The application shows the
initialisation bar, reaches the end and stops. When I revert back the
changes to check for 7.0.14, it works fine!
What is happening? How can a check for player version stop my application
from loading? I have installed 8.0.22 which I double checked on the
macromedia website! This happens on both Firefox and IE browsers.
Any help is appreciated,
Kam.
--
Flexcoders Mailing List
FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com
Yahoo! Groups Links
* To visit your group on the web, go to:
http://groups.yahoo.com/group/flexcoders/
* To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
* Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
[flexcoders] Re: How to handle manipulation of Date object and Timezones elegantly in Flex?
Hi Jim, Your problem appears to be half the problem that I am having. In your case, the client requests (for record insertion of a date) that is sent back to the server means that the server tries to translate the time into it's local timezone. Our problem inlcudes the above, plus any date object that the server sends to the client also has the same problem. The client translates the server sent date object into it's local timezone. In any case, the solutions are: 1. To do offsetting at the client end 2. To do offsetting at the server end 3. To use custom defined object for date/time without any timezones. #1 and #2 can be tricky if you have many places where the date object is referenced. For example, you may miss some date references, you need to check that every date object really does need offsetting, there is overhead when doing offsetting, you need to test thoroughly through all timezones just in case you there are some quirky side effects. #3 may require significant refactoring of existing methods. Some of the underlying issues are with the implementation of Date object: 1. In Flex (ActionScript), the Date object doesn't allow manual setting of the timezone 2. In Java, the Date object is deprecated and harder to manage than a Calendar object. So as far as I know, there isn't a way of ignoring timezones when using the Date object. I assume that using remote objects and passing of the Flex Date object as a parameter, we can only map the the Java Date object to it. It would be nice to map a Java Calendar object instead. Cheers, Kam --- In flexcoders@yahoogroups.com, Jim Laing [EMAIL PROTECTED] wrote: On 12/22/05, Matt Chotin [EMAIL PROTECTED] wrote: Are there other developers out here who've been building apps that need to work in multiple timezones (or store their dates in GMT but work in another)? I know I've answered some issues before in the past, if anyone wants to speak up ;-) We have a similar but slightly different use-case (and unresolved problem, for that matter): Our clients will be all over the world and will be sending dates back to our database via RemoteObject calls. When the date object is serialized, Flex automatically takes care of updating the time zone information. However, this is actually *not* what we want to happen because we are sending back dates with 00:00:00 timestamps (i.e. no time information). However, when this information crosses timezones during serialization, 00:00:00 becomes a time (midnight), which is then offset according to the timezone difference, and the date possibly becomes a day off. E.g. a date of 25-Dec-2005 00:00:00 GMT on the client becomes 24-Dec-2005 17:00:00 EST in the database. And those of us who celebrate Christmas know that 24-Dec != 25-Dec If there was a way that we could configure our clients to work in the same timezone as our server, then we'd be all set. Alternatively, if we could configure serialization to ignore timezone information then our dates wouldn't be corrupted. Maybe there is already an easy solution. I haven't had a chance to dig into the problem too much. But any suggestions are appreciated. Jim Yahoo! Groups Sponsor ~-- Get Bzzzy! (real tools to help you find a job). Welcome to the Sweet Life. http://us.click.yahoo.com/KIlPFB/vlQLAA/TtwFAA/nhFolB/TM ~- -- Flexcoders Mailing List FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com Yahoo! Groups Links * To visit your group on the web, go to: http://groups.yahoo.com/group/flexcoders/ * To unsubscribe from this group, send an email to: [EMAIL PROTECTED] * Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
[flexcoders] Re: How to handle manipulation of Date object and Timezones elegantly in Flex?
Dear Matt, Thanks for the link. I've added a new wish to the list. We have played with the getTimezoneOffset and have been using it. It is just not the most elegant solution, especially when the application has many date objects for both display and as part of the parameters used for requesting data from the backend. I wondered if it would be worth considering creating a custom object (or even passing a string denoting the various date elements) to pass around date and time without any timezone information, if we are very sure that we won't use timezones in the future... Kam. --- In flexcoders@yahoogroups.com, Matt Chotin [EMAIL PROTECTED] wrote: You can file enhancement requests like the formatter with timezone support at http://www.macromedia.com/go/wish. I assume you've played with setting the minutes of the date based on the getTimezoneOffset returned from the date? Matt Yahoo! Groups Sponsor ~-- Most low income homes are not online. Make a difference this holiday season! http://us.click.yahoo.com/5UeCyC/BWHMAA/TtwFAA/nhFolB/TM ~- -- Flexcoders Mailing List FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com Yahoo! Groups Links * To visit your group on the web, go to: http://groups.yahoo.com/group/flexcoders/ * To unsubscribe from this group, send an email to: [EMAIL PROTECTED] * Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
[flexcoders] Flex Enterprise Server 2.0
Hi all, I was at the Macromedia Max 2005 in Hong Kong and got a 'taste' of the new Enterprise Services in 2.0. One of the interesting things is the messaging service and the ability to push data to the client. The issue I have is that if it is using the traditional sense of publishing data to be listened to, how does such a feature deal with potentially large amount of data such as the stock market information? In a 'real time' scenario, stock market information is being published at very short intervals and transactions can be heavy depending on market and specific stocks. On the face of it, this seems like the best scenario for such a messaging feature. If we publish all stock information, then it seems like the client will have to manage the data that it wants to display. This could be expensive for the client. Maybe we divide the channels up so that different information (different markets) are published on different channels? This could reduce some of the workload for the client. In any case, if a market has thousands of stocks, then the client would still need to do a lot of expensive processing. The question is: Is there someway for the publisher to recognise different listeners and publish accordingly? Theorectically, this should be an acceptable way of publishing data... But can someone tell me if this would be possible with the upcoming Enterprise Services? Or does this all depend on the implementation of the messaging system (e.g. JMS, MQ etc)? Thanks, Kam. Yahoo! Groups Sponsor ~-- Get Bzzzy! (real tools to help you find a job). Welcome to the Sweet Life. http://us.click.yahoo.com/KIlPFB/vlQLAA/TtwFAA/nhFolB/TM ~- -- Flexcoders Mailing List FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com Yahoo! Groups Links * To visit your group on the web, go to: http://groups.yahoo.com/group/flexcoders/ * To unsubscribe from this group, send an email to: [EMAIL PROTECTED] * Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
[flexcoders] Session Management using Flex
Hi all, I'm new to Flex so bear with me. We have an existing application that we are adding user management to (login process, authentication etc). We want the capability of disallowing multiple users from logging on with the same username/password. Ideal situation is that when one user has logged in, no other user is allow to log in with the same username/password. Of course this would mean that the system would need to know when the first user has logged out (otherwise the user would never be able to log in again). A second option is when the second user logs in, the first user will be disallowed to continue as the system has checked that the first user has an invalid session. In this case, on every backend request, the system has to authenticate the user and make sure the user is logged in only once. So the questions: 1. Can the system ever know when a user has logged out in a browser environment where the user can easily close the browser without going through some sort of logging out process? 2. If we implement the second option, instead of doing a major overhaul of every method call and adding an extra parameter for the user details (e.g. username, password, session id etc), is there an existing flex functionality that provide some sort of session id that we can check which will allow us to see if the request is made from the 1st user or the 2nd user? 3. Does the AMF gateway allow some sort of session management that we can leverage for disallowing multiple users logging in with same username/password without refactoring all the backend request methods? Any help would be much appreciated. Kam. Yahoo! Groups Sponsor ~-- Fair play? Video games influencing politics. Click and talk back! http://us.click.yahoo.com/2jUsvC/tzNLAA/TtwFAA/nhFolB/TM ~- -- Flexcoders Mailing List FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com Yahoo! Groups Links * To visit your group on the web, go to: http://groups.yahoo.com/group/flexcoders/ * To unsubscribe from this group, send an email to: [EMAIL PROTECTED] * Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
[flexcoders] Re: Session Management using Flex
Dave, Thanks! Any other option other than using a J2EE application server for implementing JAAS for security? Are there existing features in Flex that we can utilise? Our application up to now has been quite simple in its deployment using a tomcat server. We're up against time in getting it into production, and deployment into a J2EE server at this stage may not be right for us at the moment. I agree with you that utilising J2EE would be simpler. Would there be a significant overhead migrating a previosly tomcat deployment to a J2EE application server such as JBoss? Issues with performance etc? We're not utilising EJB in our biz logic. The application is mainly alot of reads from the database and some calculations. Thanks, Kam. --- In flexcoders@yahoogroups.com, Dave Wolf [EMAIL PROTECTED] wrote: First off, I strongly reccomend not trying to roll your own security solutions. J2EE includes a very robust and well tested security model. Every J2EE server supports pretty much the same level of security services. There is no need to go out and develop your own. Flex integrates into the normal J2EE session security extremely well. You requirements are a bit different from the out of the box J2EE security but can be easily implemented via a JAAS plug-in extension to the container. So the questions: 1. Can the system ever know when a user has logged out in a browser environment where the user can easily close the browser without going through some sort of logging out process? Yes and no. You cannot tell (easily) when someone closes their browser but you can tell when their credentials are no longer valid. There are events which fire when a J2EE session expires. You could detect that and remove them from a list of logged in users. 2. If we implement the second option, instead of doing a major overhaul of every method call and adding an extra parameter for the user details (e.g. username, password, session id etc), is there an existing flex functionality that provide some sort of session id that we can check which will allow us to see if the request is made from the 1st user or the 2nd user? I cannot find a way to express strongly enough that the idea of adding parameters to each method call to pass security credentials is a security whole as big as a fire truck. This allows a man in the middle to very easily hijack another users session and simulate their login, thereby doing things like executing transactions by pretending to be someone else. If you use standard J2EE session based security flex will transparantly inherit the normal J2EE session. Here is how I would build this. Create a custom JAAS plugin that tracks active logins via some map. When a user logs in you add their id to the map. When they log out, you remove it. When their session expires automatically you remove it. (yes there will be lag there). If a user tries to login twice, tell the JAAS plugin to deny them. 3. Does the AMF gateway allow some sort of session management that we can leverage for disallowing multiple users logging in with same username/password without refactoring all the backend request methods? Let the container do this. This solution works identially over all data access layers as well. Flex just integrates so beautifully with J2EE security the last thing you want to do is try to roll your own security, especially by passing credentials all over. -- Dave Wolf Cynergy Systems, Inc. Macromedia Flex Alliance Partner http://www.cynergysystems.com Email: [EMAIL PROTECTED] Office: 866-CYNERGY Any help would be much appreciated. Kam. Yahoo! Groups Sponsor ~-- Most low income households are not online. Help bridge the digital divide today! http://us.click.yahoo.com/I258zB/QnQLAA/TtwFAA/nhFolB/TM ~- -- Flexcoders Mailing List FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com Yahoo! Groups Links * To visit your group on the web, go to: http://groups.yahoo.com/group/flexcoders/ * To unsubscribe from this group, send an email to: [EMAIL PROTECTED] * Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
