Simpler alternative to codebase server.

Hello people.
I understand that one needs a code base server to be a repository for code that might be downloaded so to be executed on the machine which has downloaded it.
I have alluded to this in earlier posts; I desire an alternative to this code base server way of handling code distribution. What I would like to do is remove the necessity of having this code base server third party, so a client only has to consult the same service it gets everything else from.
Any suggestions? I'm reasonably thick-skinned, and won't be hurt if they make me look a bit ignorant. I will be thankful for the enlightenment.

ejp wrote:
Additionally, my server application will deploy the code to a client which is running a client application (not a web browser) that can talk to my server.Well it won't unless you find a solution that will do that. RMI code mobility and JWS are the only solutions I'm aware of. Both require a Web server, i.e. in practical terms IIS, Apache Web Server, Tomcat, ...
The code that runs on the client needs to exchange information with the server from where it was downloaded.So it all depends on what exactly you mean by 'server'. If you mean 'host', JWS and RMI code mobility do that. If you mean XXX for some other XXX, I'm not aware of any other XXX that will do that. If you mean XXX = yur own server code, you'll have to write it, and there's no point to that.Yes, I was a bit ambiguous. Hopefully this clarifies everything.
My client application needs some code from elsewhere to complete its function. This 'function' relates to how it must exchange information with a server, and I intend this 'elsewhere' to be from a server to which the client must first connect to.
Java has this thing in RMI called a Codebase Server. One might think I'd be happy using a third party server to store code. Alas; code is data. One should be able to use RMI to get code from one program to another as if it were any other data.
So why not code?
Excuse me for seeming presumptuous, but this looks rather like a lapse in simple thinking. It hence appears to imply that I have to do this "simple thinking" myself. However, its absence may mean that I will have to delve into the innards of RMI and Java class loading. This will be both a distraction and a delay in delivery of a working application. If only simple thinking were practised more often.
If there is anyone who has had this simple thought before me, I would welcome your help.

Similar Messages

Not to reguest classes from the codebase server if not found in archive?

i am not sure but i recall since 1.6 there is some applet tag parameter to switch off class/resource retrival from the codebase server after looking in the archive jar...
i.e. i do not want my browser to flood my www server with small reguests since the appled in fact does not need anything but the jar file content...
what is this parameter?

i've recovered it :)
<PARAM name="codebase_lookup" value="false">

Simple HTTP Receiver - HTTP server code 405 reason Not Allowed explanation

Hi, I am using simple HTTP Receiver to receive the XML file from the below url http://www.bank-ua.com/export/currrate.xml
But receiving the below error in SXMB_MONI
<SAP:Code area="PLAINHTTP_ADAPTER">ATTRIBUTE_SERVER</SAP:Code>
<SAP:P1>405</SAP:P1>
<SAP:P2>Not Allowed</SAP:P2>
<SAP:P3><html> <head><title>405 Not Allowed</title></head> <body bgcolor="white"> <center><h1>405 Not Allowed</h1></center> <hr><center>nginx/1.4.1</center> </body> </html></SAP:P3>
<SAP:P4 />
<SAP:AdditionalText />
<SAP:Stack>HTTP server code 405 reason Not Allowed explanation <html> <head><title>405 Not Allowed</title></head> <body bgcolor="white"> <center><h1>405 Not Allowed</h1></center> <hr><center>nginx/1.4.1</center> </body> </html></SAP:Stack>
<SAP:Retry>M</SAP:Retry>
</SAP:Error>
Initially the same URL was working could you advice what to change to make this working?
Thanks,
Dhill

Hi Senthil,
There is a way, you can expose the function module HTTP_GET as an rfc-enabled FM and use it to call the webservice. Here is a POC I made before:
Proof of Concept: An Alternative HTTP GET using a Remote-Enabled Version of the Function Module HTTP_GET
Another is to create a UDF that uses an HTTP Client class to call the webservice.
Regards,
Mark

Alternatives to Leopard Server

Given the number of glitches , bugs , Documentation fuzziness and a few more subjects we are considering dropping mac os X server has a production platform on newer systems and we are looking for Working stable alternatives to os X Server, i hope we can come back to os X server in the future but until then we will have to switch to something else.
Any suggestions will be more than welcome and studied in detail....
I am sorry if this sounds off topic in the Mac os X Server forums but since we are getting so little feedback about the many issues plaguing mac os X Server from Apple in Europe (yes even through support calls that keep mentionning "we have an update coming for the last 2 months) none of them as resolved the issues we are facing.
Thanking you for your time and attention and waiting to read from you soon ,
Best Regards,

Camelot wrote:
At the end of the day Mac OS X Server is not going to be the best tool for every job.
Server Admin does a reasonable job for common setups, but there are going to be a lot of people who step outside of the box that Server Admin lives in. I know I'm one of them. Whether that means you're no better off than you would be on another platform depends on what your priorities are. If a unified GUI is important to you and Server Admin isn't up to your needs then Mac OS X may not be the best option. If, on the other hand, you don't mind stepping away from Server Admin in the cases where you need more than it offers you still get the other benefits of OS X.
As we are in the 'web' portion of this forum, I'll only address my thoughts on the web tools. I agree with most all that Camelot and others have said. Particularly where one doesn't want to rely on one server for all services -- but even that is predicated on load over all else. Surely a 10 person company with one website, email and off-topic services such as file sharing in a mixed environment could be completely happy with a high-end mini or any Xserve for all services.
However, part of what Apple is selling is a web-capable server. At 10.5.2, OSXS is NOT a great GUI-managed tool for web services. Not only does it still have bugs with regard to Site management and choices the Server group made with configuring httpd (as well as an anemic php/mysql implementation), there are significant limits to using services such as ftp, mail, even listserves.
I remember thinking when they hired the likes of Eric Z from Starnine (as it was in the throws of building Webstar 4 and being sold to 4D) that maybe he would bring some fine ideas to the Server group. He wasn't technically proficient like real coders such as CJ, but he had some understanding of what Webstar had in terms of strengths. To date, I've not seen a lot of evidence that those strengths have made it to the OSXS product. They didn't in Tiger, they aren't there in Leopard 10.5.2 (yet).
Where is a simplified GUI admin for supporting multiple FTP accounts to the same home/site (or subpaths within a site)? Where is a decent GUI for admin of php4 settings, even something as simple as upping the .ini file size limits? Where is a decent GUI for supporting multiple domain names in the listserve product? Why does performance cache redirect to an alternate port when an iWeb user uploads oddly-crafted XML tags and the php interpreter parses the page incorrectly? Where is there user-assignable admin rights to email accounts for a specific domain so a sysadmin can delegate account management to a client?
The list goes on and on. I'm appreciative of what Apple has been able to do with much of the 'web' part of the GUI, but there are many, many features that can be GUI-ized (and should be) if Apple want to continue to push the paradigm that the machine should not get in the way of the user. I've a stack of 20+ year old Addison publications where Apple spent much money and time pushing the user experience concept. To have the unix folks push back and state that only CLI is the answer is -- NOT the answer. It is not how Apple should expect even sysadmins to need to go. And hey, I'm plenty comfortable myself using CLI. But my customers should not have to come to me with every small change for feature sets that can easily be added to the GUI, either under the SA, the WGM, or elsewhere in the GUI toolset.
OSXS is not really a mature Webhosting server. It is a server that takes the bare bones of Apache, of older PHP tools (why not offer updated modules or ancillary modules for PHP5?) and mySQL tools (same question). Their own scripts cannot properly or reliably migrate a 10.4 installation with Apache 1.3 to Apache 2.x in 10.5. Even the error logs are too obscure to offer much assistance tracking down the actual failures when they occur. Same has already been said of the DNS module.
I'm trying to stay in bounds here in my critique -- as an enterprise server, OSXS is coming along. It has a lot of maturation to undergo in the 'web' realm, as it does in collaborative sharing (iCal/contacts/etc.). Let's hope by continuing to purchase and deploy it, we can make enough of an impact that Apple grows the product accordingly and doesn't abandon it as they did the Xraid stagnation.

Opinion: simple alternative to DNS

There's been some nerd news recently about people being upset with ICANN and wanting an alternative DNS. (example at OSNews)
That seems like a whole lotta work to me. All DNS does is make "www.google.com" point to "74.125.230.80", right? It seems like we just need a web browser that makes working with IP address easy, for example by providing the IP address to a variety of search engines and popular websites by default.
Does the "supreme rule" of ICANN upset you? Would you use raw IP addresses instead of domain names? Is there something about this simple solution that I'm missing?

The obvious: how do you memorise an IP address like 23.17.475.9? What DNS does now is maintain a database, people might be complaining about ICANN, but if you delegate that task to e.g. Google, what guarantees do you have the results will be correct? In other words: how will that improve the already existing problems?
Afaik the main issue people seem to have is that there is a root server located in China that handles increasingly more DNS requests, even from non-Chinese computers, but it is said to censor those requests nonetheless, be they from Chinese or foreign systems.
Even though feeding the data to multiple 'agents' might allow people to catch the rotten apples easier, it wouldn't be any different to the way it is now - you can use different DNS servers now if you want to. Your ISP probably has its own DNS servers, ICANN has theirs, OpenDNS is there, I know the company behind the Comodo security suite for Windows has its own as well.

Looking for a simple photo gallery web server setup

Hi,
I have Arch running on my low powered AMD HP microserver and am looking at letting relatives,friends etc see my photos over the web.
I have php and sqlite3 already installed and dyndns setup.
What further would I need (I'm looking for as painless, light and small as possible setup)
Cheers
Smiffy

Yes - I've tried that and could try a few different setups until I find what I want.
I was more looking for any recommendations especially from someone who has set something like this up before.
This page was useful :-
http://www.yolinux.com/TUTORIALS/LinuxT … llery.html
BBGallery looks good and simple to me but needs GIMP (I have a headless server so don't really want to install extraneous sw)

OBIEE 11.1.1.6 as Simple Install , Error in Server logs

Hi Guys,
I have newly installed OBIEE 11.1.1.6 as Simple Install and I am finding this error in my server logs, So please can you guide me to resolve this impending issue?
####<Apr 23, 2012 12:06:06 PM IST> <Notice> <WebLogicServer> <Vijay-PC> <AdminServer> <Main Thread> <<WLS Kernel>> <> <> <1335162966866> <BEA-000365> <Server state changed to FAILED>
####<Apr 23, 2012 12:06:06 PM IST> <Error> <WebLogicServer> <Vijay-PC> <AdminServer> <Main Thread> <<WLS Kernel>> <> <> <1335162966866> <BEA-000383> <A critical service failed. The server will shut itself down>
####<Apr 23, 2012 12:06:06 PM IST> <Notice> <WebLogicServer> <Vijay-PC> <AdminServer> <Main Thread> <<WLS Kernel>> <> <> <1335162966898> <BEA-000365> <Server state changed to FORCE_SHUTTING_DOWN>
####<Apr 23, 2012 12:06:06 PM IST> <Info> <WebLogicServer> <Vijay-PC> <AdminServer> <Main Thread> <<WLS Kernel>> <> <> <1335162966991> <BEA-000236> <Stopping execute threads.>
Thanks,
Vj

Hi Vj,
The error logs are quite generic and do not seem to help much. If you are on windows, check the event viewer if there are any messages. You may want to check the Admin server's ".out" file too other than ".log" file.
Hope this helps.
Thank you,
Dhar

Simple dmz for mail server

If I have a normal 1700 with a serial and 2 ethernet interfaces, can someone give me an example how to put this exchange server in a dmz ? I could do with a simple config as a pointer
thanks guys

Carl,
Here's a pointer to get you started.
http://www.cisco.com/en/US/products/sw/secursw/ps1018/products_configuration_example09186a0080094111.shtml
Let me know if it helps,

Alternative of the server restart for Model re-import

Hello Everybody,
Whenever we re-import RFC model, we need to restart WAS J2ee stack to make it effective. But, there should be the better solution for this. This could be tolerable for developement system, but this is not the way for production system as it affects so many applications running on production system.
Is there any other way to avoid system restart?
Regards,
Bhavik

Hi,
That is showing the changes in the Model i.e., Modified and Newly Added fields/structures/tables.You will get that normally when you re-import the model.
After re-importing the model you need to modify the mapping.
Thanks n Regards,
Jhansi Miryala

How to copy procedures and packages from one server to another?

Hi,
I have 439 packages and 178 procedures located in one server called MAXWELL.
I need to copy these objects to another server called TITAN.
The schema names is SPCBR in both servers.
So both servers have an instance running with this schema SPCBR.
SPCBR in TITAN server has their tables with table data which was generated by an export/import process from MAXWELL server.
However, procedures and packages were not copied by the export/import because I used the clause 'tables'.
Now, what can I do in order to equalize procedures and packages in both databases? I can't risk in damaging or duplicating data in my database located in TITAN server, ok?
Thanks,

Another simple alternative, according to AskTom (http://asktom.oracle.com/pls/ask/f?p=4950:8:::::F4950_P8_DISPLAYID:454220160386)
---------------- getcode.sql ----------------------------
set feedback off
set heading off
set termout off
set linesize 1000
set trimspool on
set verify off
spool &1..sql
prompt set define off
select decode( type||'-'||to_char(line,'fm99999'),
'PACKAGE BODY-1', '/'||chr(10),
null) ||
decode(line,1,'create or replace ', '' ) ||
text text
from user_source
where name = upper('&&1')
order by type, line;
prompt /
prompt set define on
spool off
set feedback on
set heading on
set termout on
set linesize 100
------------------- eof --------------------------------
it extracts one procedure, function or package to a file. Thats it. If you
wanted to get all of the procedures in a schema extracted to the current working
directory, you would run a script:
--------------- getallcode ---------------------------
set termout off
set heading off
set feedback off
set linesize 50
spool xtmpx.sql
select '@getcode ' || object_name
from user_objects
where object_type in ( 'PROCEDURE', 'FUNCTION', 'PACKAGE' )
spool off
spool getallcode_INSTALL.sql
select '@' || object_name
from user_objects
where object_type in ( 'PROCEDURE', 'FUNCTION', 'PACKAGE' )
spool off
set heading on
set feedback on
set linesize 130
set termout on
@xtmpx.sql
---------------- eof ---------------------------------
You can see how to filter on that one by adding to the where clause if you want.
Just run @getallcode_INSTALL to run all of the scripts...

Welogic 9.2 jdbc connection pool monitoring section not displaying server

Welogic 9.2 managed jdbc connection pool monitoring section not displaying server name. Can you please let me know how can i achive this

Hi,
It happens if the Connectivity is Lost b/w DataBase & WLS Server...In that case in the WLS AdminConsole---DataSource---Monitoring Tab you wont see even the Testing Page with valid datasource to test it.
So please make sure that the N/W connectivity proper....One more thing you can try:
<h3><font color=red>Test-1</font></h3>
You can use *“weblogic.Admin”* utility to Enable and Disable the Pool (DataSource) Just to Confirm whether the dataSource is actually active or not::
Step1). run *"setWLSEnv.sh"* first in the same Shell prompt....then do the following:
Step2). To suspend:
java weblogic.Admin -url t3://localhost:7001 -username weblogic -password weblogic SUSPEND_POOL YourDataSourceName
Step3). To re-enable:
java weblogic.Admin -url t3://localhost:7001 -username weblogic -password weblogic RESUME_POOL YourDataSourceName
To test whether you are getting any Exception or Error while doing this....If yes u are getting any Error or exception then It means there May be some Database connectivity issue...
<h3><font color=red>Test-2</font></h3>
Please use the following Utility to Test the Database connectivity from your WLS Server box:
<font color=maroon>
java -classpath /opt/app/bea/weblogic92mp2/weblogic92/server/lib/weblogic.jar -Dbea.home=/opt/app/bea/weblogic92mp2 utils.dbping ORACLE_THIN DbUserName DBPassword somethingDbUrl.Something.net:1521:dbInstance
</font>
Syntax:
java -classpath /opt/app/bea/weblogic92mp2/weblogic92/server/lib/weblogic.jar -Dbea.home=/opt/app/bea/weblogic92mp2 utils.dbping ORACLE_THIN <DBUserName> <DBPassword> <DBURL>
The above utilty will test the DB Connectivity ...you need to run it from your WLS server Box...
<h3><font color=red>Another Simple Alternative of Test-2</font></h3>
http://jaysensharma.wordpress.com/2010/03/23/testing-datasource-status-using-wlst/
Thanks
Jay SenSharma
http://jaysensharma.wordpress.com (WebLogic Wonders Are Here)

SQL Server Tutorials for beginners: OLAP / Data Mining

I teach a DBMS + BI course to non-CS (business) students, using SQL Server. To illustrate OLAP I use the nice interactive online demo at "olaponline.radar-soft.com".
Is there a simple tutorial for SQL Server that is similar? Prepackaged, Illustrating OLAP, BI, or Data Mining algorithms? Everything I've seen so far is too complicated and requires many preparation steps before interacting with the model/cube.
Alternatively can you recommend other online tools / demos? Thanks,
-- Shaul

Hello,
The best way to learn about data mining is the list of 22 lessons created y Daniel Calbimonte:
http://www.sqlservercentral.com/Authors/Articles/Daniel_Calbimonte/1486684/
About Business Intelligence, please take the following free training:
http://www.microsoftvirtualacademy.com/training-topics/business_intelligence_topic_page_en
Hope this helps.
Regards,
Alberto Morillo
SQLCoffee.com

RMI codebase confusion

You'll have to excuse me if I am asking stupid questions, but I can't find the information I want in the documentation. I am working on an RMI client-server project with 3 jars client.jar, common.jar, server.jar. The common jar contains a Remote interface which the client looks up, and the server implements. The stub class is contained in the server jar (obviously).
Currently our method of distribution is to distribute all 3 jars to the client machines - I want to avoid this as it is unnecessary. I would therefore like to set java.rmi.codebase property in the server so that clients can download the stub classes at run-time. OK.
Our RMI app is being distributed on an internal network - suppose on the server machine (call it "ourserver.london.uk.ourco.com", the relevant Jar file would be at /usr/bin/my_app/jars/server.jar (or wherever).
The clients lookup rmi://ourserver.london.uk.ourco.com:1099/MY_EXPORTED_REMOTE. Questions:
1. Is this RMI "system" dependent on the server having an Http server?
2. If it isn't, how do I define the codebase other than "http://..."?
3. If it is (and the machine certainly does have an http server on it), how do I specify the codebase URL for the clients, given that the server is not "under" the http server (IYSWIM)?
I'd be glad of answers to these (reasonably simple) questions.
Chris

First of all, I would arrange things a bit differently. In both of the cases below, we assume a http server, it's location is unimportent, but it has to be reachable from the the client. It could be the same host as the server, or a third machine.
Option #1:
client.jar -- All client class files, including all class files of remote interfaces called by the client.
server.jar -- All server class files, including all stubs, and remote interfaces implemented and their implementations.
server_dl.jar -- Just the stubs.
The client host would contain just client.jar in the classpath.
The server host would contain server.jar and server_dl.jar in its classpath
server_dl.jar would also be placed on the http server and it's url would be the server's codebase.
Option #2:
client.jar -- All client class files, but none of the remote interfaces.
server.jar -- All server class files, but none of the stubs and no remote interfaces.
interfaces.jar -- The remote interfaces shared by client and server.
server_dl.jar -- just the stubs.
The client host would need client.jar and interfaces.jar in its classpath.
The server host would need server.jar, interfaces.jar and server_dl.jar in its classpath.
server_dl.jar would also be placed on the http server and it's url would be the server's codebase.
Now to answer your questions.
1. Is this RMI "system" dependent on the server having an Http server?If you want to have the client download the server stubs, then they have to be on an http server (I think a ftp url for the codebase may also work, but I have not tried it).
2. If it isn't, how do I define the codebase other than "http://..."?
You may be able to use a ftp url in which case the url would be "ftp://....."
3. If it is (and the machine certainly does have an http server on it), how do I > specify the codebase URL for the clients, given that the server is not "under"
the http server (IYSWIM)?The downloaded class files do not have to be on a http server that is on the same host as the RMI server. The RMI server's codebase can be any valid url (reachable from the clients) The only requirement is that the class files be reachable via the codebase. You do have to configure the client's network security so that it has permission to open connections to the codebase server.

App. server url

Simple question, but coudn't find the answer..
How can a webstart app. find out it's jnlp url ?
In fact, I am trying, in a webstart "main app", to run a webstart "help app",
so I want to exec() "javaws the_Child_Url_constructed_From_The_Main_Url".
Also, I have seen somewhere that giving the url as arg to javaws will only work
in access to the server is possible (server not down for ex) and to avoid that
one has to "reconstruct" a file:/ url pointing to the .javaws cache....
Any hint is welcome :-)

import javax.jnlp.*;
private BasicService m_basic = null;
try {
     m_basic = (BasicService)ServiceManager.lookup("javax.jnlp.BasicService");
} catch (UnavailableServiceException use) {
         System.err.println(use.toString());
         return;
m_basic.getCodeBase().toString(); // Return codebase server URLThe javax.jnlp API classes can be found in the javaws.jar file.

Windows Client log to the Mac Os X Server ?

Hello everybody
I would like to create a network with Mac OS X Server. I have Windows PC's that would also be in the network.
Does windows clients can connect to the Mac Server with a username and password ?
Thank you.

At the simplest, sure, Microsoft Windows clients can connect to OS X Server services with appropriate credentials. Mail and file sharing, for instance.
Beyond password-based authentication, a more advanced configuration is probably not as simple nor as easy as you might want. Windows clients really want to use Microsoft Active Directory here and Kerberos authentication probably won't get you where you want. (This assuming that you can get the Windows clients to accept OS X Server Kerberos authentication.)
The usual approach for single sign-on in heterogeneous environments is the so-called Magic Triangle configuration with both Microsoft Active Directory running on a Windows Server and Open Directory running on OS X Server. This involves the two domains communicating. It might be easier to get OS X clients operating off Windows Server, particularly with the use of profiles in more recent versions.
As an alternative to Windows Server, recent versions of the Samba server package include authentication services compatible with Active Directory, and you may be able to get your Windows clients to authenticate to that, and build that as part of a Magic Triangle configuration. You'd need to download and configure the Samba server, and get the Samba AD support working.
Related discussions and documentation here and here and maybe this thread. There are other discussions.

Simpler alternative to codebase server.

Similar Messages

Maybe you are looking for