Connecting two Coherence Clusters using Replication

I am using: http://wiki.tangosol.com/pages/viewpage.action?pageId=1343616 to setup the "Replicated-site Coherence*Extend Example". On running the build I am getting errors as shown below. Can anyone help me to resolve this so that I can get the example up and running?
.\..\src\com\tangosol\examples\extend\ExtendReadWriteBackingMap.java:57: cannot find symbol
symbol : constructor ReadWriteBackingMap(com.tangosol.net.BackingMapManagerContext,com.tangosol.util.ObservableMap,java.util.Map,com.tangosol.net.cache.CacheStore,boolean,int)
location: class com.tangosol.net.cache.ReadWriteBackingMap
super(ctxService, mapInternal, mapMisses, store, fReadOnly,
^
.\..\src\com\tangosol\examples\extend\ExtendReadWriteBackingMap.java:85: incompatible types
found : com.tangosol.net.cache.ReadWriteBackingMap.CacheStoreWrapper
required: com.tangosol.net.cache.CacheStore
CacheStore store = getCacheStore();
^
.\..\src\com\tangosol\examples\extend\ExtendReadWriteBackingMap.java:153: storeAllInternal(java.util.Set) in com.tangosol.net.cache.ReadWriteBackingMap.CacheStoreWrapper cannot be applied to (java.util.Map)
super.storeAllInternal(mapEntries);
^
3 errors
On running javac in verbose mode it does show that the classes in coherence.jar have been loaded, especially the one required:
[loading com\tangosol\net\cache\ReadWriteBackingMap.class(com\tangosol\net\cache:ReadWriteBackingMap.class)]
[loading com\tangosol\net\cache\ReadWriteBackingMap$CacheStoreWrapper.class(com\tangosol\net\cache:ReadWriteBackingMap$CacheStoreWrapper.class)]

I am running with Java: jdk1.6.0_21, Coherence: 3.6 and ant: 1.8. Nb. I am running on Windows but that should not matter....
I thought it could be a problem with ant not passing in the correct classpaths or somehting similar so I wrote the equivalent DOS script (shown below) to run the build but still got the same error.
Nb. I have managed to run the synchronized site example with the above java, coherence and ant versions.
build.cmd_
@echo off
if "%1"=="build" goto build
if "%1"=="clean" goto clean
goto help
:help
echo "usage: build.cmd build"
echo build: makes the clases directory and compiles the java
echo clean: removes the classes directory
goto exit
:build
mkdir .\..\classes
"%JAVA_HOME%"\bin\javac -d .\..\classes -cp "%COHERENCE_HOME%"\lib\coherence.jar .\..\src\com\tangosol\examples\extend\*.java -verbose
goto exit
:clean
rmdir .\..\classes /q/s
goto exit
:exit
"%JAVA_HOME%" = "C:\Program Files\java\jdk1.6.0_21"
"%COHERENCE_HOME%" = "C:\Program Files\Oracle\coherence-java-3.6.0.0b17229\coherence"
Edited by: Ku on Sep 21, 2010 11:37 AM
Added info to show location of java/coherence variables.

Similar Messages

Connecting Two PC's Using Minimum Hardware !

What are the possbile ways to connect two PC's using minimum hardware ?

What is single flat layer 2 switch ?
(Cross over cable You mean Lan Cable)
========================================
I have two PC's which I want to use in a Client-Server setup.
(both pc's have lan card, usb ports)
(Qus 1). What Minimum Set of HW I require ?
(Qus 2). What Minimum Set of SW I require ?

Connecting to multiple clusters using WAR file

I am trying to connect to multiple clusters by creating separate WAR(web archive) files through Apache tomcat.
My intention is to have each 'war' act as a node to a separate cluster. I started with the 'jmx-console' example available on the Tangosol website.
So, my understanding is that even though all the war files are loaded in the same JVM (tomcat), since each of these war's has its own classloader, we can have multiple tangosol ndoes from the same JVM.
So I have two basic questions:
*) If I use system.setProperty("tangosol.coherence.override", myOverrideFile-x) in the class loader of each WAR, can I have the nodes connect to different clusters. is'nt the System.setProperty() setting the property on the JVM? If so, how can multiple classloaders running in the same JVM connect to different clusters?
*) Where should I place config/override xml files under the tomcat directory structure for it to be accesible to the class loader.

Hi sumax,
Thanks for your response Robert,
Maybe I did not understand your solution fully, I
thought your idea mostly talks about how to configure
the override files, and not how to configure the
individual nodes to use different override files.
That part is taken care of by the fact, that each war file loads the tangosol-coherence-override.xml from its own WEB-INF/classes directory.
Classloading from a web app in Tomcat resolves classpath classes/resources in the following order:
1. java.* and javax.* and com.sun.* packages are loaded from the system classloader. This, I believe, cannot be circumvented on a Sun JVM with a classloader extending a JDK classloader class.
2. Reloadable JSP classloader (loads servlet classes generated from JSP pages). This is webapp specific, each webapp sees its own generated servlet classes.
3. Reloadable Webapp classloader (loads from WEB-INF/classes and WEB-INF/lib of the war). This is webapp specific, each webapp sees its own WEB-INF/classes and WEB-INF/lib.
4. Common classloader (loads from ${TOMCAT_HOME}/common/classes and ${TOMCAT_HOME}/common/lib). This is used by all webapps and the Tomcat container itself. You should not put anything here, except for libraries for resource drivers, etc. JDBC drivers or anything which needs to be published in the JNDI server in a non-application-specific way.
5. Shared classloader (loads from ${TOMCAT_HOME}/shared/classes and ${TOMCAT_HOME}/shared/lib). This is shared between all webapps but not by the Tomcat container. You can put libraries here with which you want to communicate between webapps. Any singletons in classes in this classloader will be shared between webapps. A place for configuration files which are shared or if they have the application name in their filenames (so that they are not accidentally used by another application).
The question still lingering in my mind is:
How do I tell the classloader which override file to
load.
A tangosol-coherence-override.xml with different content should be put in the WEB-INF/classes directory of each war file. Each of those files will be loaded only by its containing webapp. This filename is defined in the deployment-mode-specific override files in coherence.jar or tangosol.jar (I don't remember which, off my head) which all chain to this same tangosol-coherence-override.xml file.
That override file can contain all your overrides, if you do not want to change those between restarts, or it can chain with its xml-override attribute (as shown in the example) to another file with the application name in its filename.
This chaining allows each webapp to refer to a differently named override file, so all those override files can be put into shared/classes folder. The uniquely named override file in shared/classes allows you to change the override configuration between restarts.
As for cluster address and port, you can specify those in any of the override files.
The full chaining path is the following (if you used the tangosol-coherence-override.xml example in my first reply):
tangosol-coherence.xml -> tangosol-coherence-prod.xml -> tangosol-coherence.xml (you provide this file in your webapp WEB-INF/classes) -> tangosol-coherence-override-warfilename.xml (you provide this file in the ${TOMCAT_HOME}/shared/classes)
I hope this makes it clear.
Best regards,
Robert

Connecting two Mac Minis using wifi

My actual set up:
Bedroom - Macmini (2007) linked to a dsl modem via an ethernet cable. A Samsung display, a LaCie 2tb drive, are also connected to it.
In a very near future:
LivingRoom - Macmini (2011, to be delivered in a couple of days... hopefully) will have access to internet via wifi. A 55" TV, a LaCie Minimus 2tb drive (a clone of the other one in the bedroom) and a DAC will be connected to the new mini.
My question:
How can I connect the two minis together using wifi so that I can have easy acces to all the files in the drive connected to the Mac mini in the other room?
I have looked at the "File sharing preferences" in the system preferences but I am not that sure how to do it.
A little guidance would be greatly appreciated!

"File sharing preferences" is indeed how to do it. Connect both minis to your router (one by ethernet cable, one by wifi, it doesn't matter.)
Then turn on "File Sharing" in Sharing Prefs on both minis. Also, give them both meaningful, different names under "Computer Name".
Then, when you open a finder window on one mini, the other one should appear under "Shared", like this:

Connecting two computers, wirelessly, using Airport?

I have a IMac intel and a PowerBook G4 - 400 Ti. They both have Airport cards. I understand that they will "talk" to each other with out benefit of a router. I would like to know how I can set this up. I have gone through the setup procedures endlessly and, although I did manage to get the IMac to see the PB once, and I don't know how it happened, I have not been able to do it again. There was an icon on my IMac desktop that I could drag stuff to and it would transfer. I'm quite confident its a simple procedure that I, in my stupidity, cannot fathom. When I try to set up a network I get an error message saying : "There was an error joining the Airport Network". Could it be a "slave/master" relationship between the two computers? With a proper setup will the flow of info be two way or one way? Are the two computers or systems incompatible? You help will be deeply appreciated!
John Callahan

To my knowledge the information you heard is not
correct.
http://docs.info.apple.com/article.html?path=Mac/10.4/
en/mh914.html
Any computer that has an Airport Card, or Wireless
Card (PC) will recognise any network that they are in
range of by and with the use of a ROUTER, thus
completeing the 2 way communication.
There are several different ways to connect and
confiqure a network, these different ways can be and
can get confusing without the paperwork on the step
by step set-up.
http://compnetworking.about.com/cs/basicnetworking/f/w
hatsnetworking.htm
Cheers Don
Actually none of the Airport cards require a Wireless base station (acting as a router/access point) to work. Airport supports ad-hoc wireless mode (computer-to-computer mode, in Apple lingo), which allows two or more wireless-enabled computers to be connected to each other WITHOUT the use of base station. Search for "computer-to-computer" in the support knowledge base.

Tecra A9 - Connecting two external displays using Express Port Replicator

Hi,
I have a Tecra A9 and Express Port Replicator.
I'm trying to connect 2 external displays, one connected on the notebook output and the other connected on the docking station (notebook lid closed, not using the notebook display).
I can only get one to work, does anyone know if this is possible?

Hi
I think you want to connect one monitor to VGA and the other to DVI port.
Is it right?
Well, I found this doc:
http://aps2.toshiba-tro.de/kb0/TSB7B0201000CR01.htm
It depends on the used Tecra A9, if you can get a picture on the DVI out port on the Express Port Replicator (PA3508E-1PRP) or not.
The DVI port on the on the Express Port Replicator is only supported on machines with ATI or nVidia graphic card.
If you have a INTEL graphic card the DVI port on the on the Express Port Replicator is NOT supported.

Connecting two computers together using an intermediate server.

I'm trying to write a program that allows users to connect to a server and then pairs off users to chat. I setup the connections between the two clients and that works fine, but as soon as the server disconnects from the two clients the connection between the two is lost. The server is always runnning, it sets up sockets that are global to the clients, what am I missing?

I am not very sure how are you trying to do the job. Is it
client1 <-----> server <----> client2
i.e. all message from client1 is send to server which passes them on to client2 and vice versa. If yes, you will loose connectivity when server goes down. Instead, try a an connection client1 directlyt to client2. for this you might need to run a socket server on client2
the pseudo code might look like this
1. client1 connect to server.
2. client1 check client2 is logged in and convays interest to connect to client2
3. client1 gets IP address and port of client2 from server. This info has been convayed to sever when client 2 logged in.
4. client1 opens socket connection to client2
5. client2 has ServerSocket with establishes recieves connection, opens socket and starts chat seesion.
Note that after step 5, client1 and client2 are directly connected and server no longer comes into picture. It only acts as repository for IP address and port of client2.
hope it helps
Amitabh

Connecting two 21" imac displays

Is it possible to connect two Imacs and use the displays as one? If so, someone please let me know what I need to do...
Thanks

Hello, which iMacs are these?
So we know more about them...
At the Apple Icon at top left>About this Mac, then click on More Info, then click on Hardware> and report this upto but not including the Serial#...
Hardware Overview:
Model Name: iMac
Model Identifier: iMac7,1
Processor Name: Intel Core 2 Duo
Processor Speed: 2.4 GHz
Number Of Processors: 1
Total Number Of Cores: 2
L2 Cache: 4 MB
Memory: 6 GB
Bus Speed: 800 MHz
Boot ROM Version: IM71.007A.B03
SMC Version (system): 1.21f4
If nothing else...
ScreenRecycler...
http://www.screenrecycler.com/ScreenRecycler.html

Coherence clustering between two separate wka clusters

Hi,
I am running Coherence 3.6 in my environment.Off lately I am seeing a weird issue with respect to two separate wka clusters. when I mean separate wka cluster, these are two clusters configured with a distinct wka list.
I try to connect with Coherence cmd line tool to a cluster named "cluster-ABC" which has a wka IP- 10.30.xx.xxx on "host x". Then I come out of this cluster, and connect from the same command prompt window to another host -"host y" which has a same named cluster-"cluster-ABC" but a different wka IP-10.30.yy.yyy. Between the two clusters, the only common factor is the cluster name. Other things such as wka list etc are different.
After some time, I see that nodes from "host y" start joining "host x". this seems to happen only when i connect to two different clusters, from my command line utility. The clusters here are formed using wka in each case. there is no multicast enabled on our setups. This used to never happen with coherence 3.5, where i did similar kind of operations on the cluster.
Is there some kind of reference being held by the utility even after it is disconnected from the second cluster?
Thanks,
Chandini

We tried out your suggestion on using unique cluster names. I had 2 hosts each running a cluster with a unique cluster name; cluster-A and cluster-B. I had another node on a third host connect to cluster-A using port 8092 and then tried connecting to the cluster-B after changing the cluster name appropriately. But, this time the node was rejected by cluster-A saying that the configurations are different.
On changing the port from 8092 to 8090 and connecting to cluster-B worked just fine. I guess, just having unique cluster names isn't helping. Would this be fixed by the ticket that's been raised?
Is there anything else we could do in the interim to get around the issue? Using different ports might not be a very helpful solution as we've noticed that 3.6 onwards, we need to open up the ports in the firewall of the host where the nodes are running. This wouldn't be too feasible for us.
-Roshan

Is it possible to connect to two different clusters?

Hi,
I was wondering if it was possible to connect to two different clusters from the same process?
What I would like to do is to able to get some cache on the first cluster and others on a second cluster. The cache-config DTD doesn't seem to forbid the definitions of several remote-cache-scheme. So I should be able to map all "c1-*" caches to the first cluster and all "c1-*" caches to a second cluster...
Is it possible?

Thanks,
Yes I'm connection via extend.
Could you give me an example of such configuration? I've made a test from a C++ client with the following client config:
====================================================================
<?xml version="1.0"?>
<!DOCTYPE cache-config SYSTEM "cache-config.dtd">
<cache-config>
<caching-scheme-mapping>
<cache-mapping>
<cache-name>c1-*</cache-name>
<scheme-name>extend-1</scheme-name>
</cache-mapping>
<cache-mapping>
<cache-name>c2-*</cache-name>
<scheme-name>extend-2</scheme-name>
</cache-mapping>
</caching-scheme-mapping>
<caching-schemes>
<near-scheme>
<scheme-name>near-c1</scheme-name>
<back-scheme>
<remote-cache-scheme>
<scheme-ref>extend-1</scheme-ref>
</remote-cache-scheme>
</back-scheme>
<invalidation-strategy>auto</invalidation-strategy>
</near-scheme>
<near-scheme>
<scheme-name>near-c2</scheme-name>
<back-scheme>
<remote-cache-scheme>
<scheme-ref>extend-2</scheme-ref>
</remote-cache-scheme>
</back-scheme>
<invalidation-strategy>auto</invalidation-strategy>
</near-scheme>
<remote-cache-scheme>
<scheme-name>extend-dist</scheme-name>
<service-name>ExtendTcpCacheService</service-name>
<initiator-config>
<tcp-initiator>
<remote-addresses>
<socket-address>
<address system-property="tangosol.coherence.proxy.address">ldnpsm020001413</address>
<port system-property="tangosol.coherence.proxy.port">4321</port>
</socket-address>
</remote-addresses>
<connect-timeout>10s</connect-timeout>
</tcp-initiator>
<outgoing-message-handler>
<request-timeout>5s</request-timeout>
</outgoing-message-handler>
<serializer>
<class-name>com.tangosol.io.pof.ConfigurablePofContext</class-name>
</serializer>
</initiator-config>
</remote-cache-scheme>
<remote-cache-scheme>
<scheme-name>extend-paul</scheme-name>
<service-name>ExtendTcpCacheService</service-name>
<initiator-config>
<tcp-initiator>
<remote-addresses>
<socket-address>
<address system-property="tangosol.coherence.proxy.address">ldnpsm020001412</address>
<port system-property="tangosol.coherence.proxy.port">4131</port>
</socket-address>
</remote-addresses>
<connect-timeout>10s</connect-timeout>
</tcp-initiator>
<outgoing-message-handler>
<request-timeout>5s</request-timeout>
</outgoing-message-handler>
<serializer>
<class-name>com.tangosol.io.pof.ConfigurablePofContext</class-name>
</serializer>
</initiator-config>
</remote-cache-scheme>
</caching-schemes>
</cache-config>
========================================================================
if I do a CacheFactory::getCache("c1-Test"), the log shows a connection to the 1st cluster
if I then do a CacheFactory::getCache("c2-Test"), the log shows a connection to the 1st cluster and not to the second cluster
What did I do wrong?

Connect two external monitors to mini DP(using a mini DP to VGA adapter) and the VGA out at once

I am trying to connect two external monitors to mini DP(using a mini DP to VGA adapter) and the VGA out at once. Windows recognize both monitors as well. But it dowsn't let me use all three monitors at once(see the below image). In the dropdown, it lists all three monitors. Problem is there are two active displays I can select.

Mavec wrote:
I would like to add a third monitor.
The GPU(s) in your T400 (and most other laptop brands) will support two different displays concurrently. If you want to support more than two different displays concurrently, you will need another GPU external to the T400.
Generally, if you want to use two external displays on a T400 you will nee dto use a docking station.
I have read messages on various fora on this topic and know that there are a number of ways of doing this, but as I have never had a need to do this myself I can't tell you how to go about it.
Cheers,
Bill
I don't work for Lenovo

Problem connecting two trunks to sip provider using same CUBE

We need to connect two SIP trunks from service provider to Cisco CUCM 7.1 using CUBE “Cisco 2821”, SP using the following configuration:
First SIP PSTN Link Configuration(In-Out DID/DOD 218 7700 – 218 7799)
Customer IP Address =   10.196.191.158/30
SP IP Address = 10.196.191.157/30
Protocol= SIP
SIP Port = 5060
Transport Protocol=UDP
Voice Codec= G711 A-Law
DTMF = IN-Band DTMF without RFC2833
Signaling IP address = 10.201.20.49
IP Address 10.201.20.10 (Media IP) must be visible from IP PABX
Second SIP PSTN Link Configuration( Inbound Only 920009999)
Customer IP Address =   10.196.192.94/30
SP IP Address = 10.196.192.93/30
Protocol= SIP
SIP Port = 5060
Transport Protocol=UDP
Voice Codec= G711 A-Law
DTMF = IN-Band DTMF without RFC2833
SIP server IP address = 10.201.20.49
IP Address 10.201.20.10 (Media IP) must be visible from IP PABX
When we tried to configure both links on the same CUBE we faced two problems:
-          Routing issue, as we can’t route traffic using single CUBE through two different interfaces to the same destination “ i.e we have to configure static route commend (ip route 10.201.20.49 255.255.255.255 10.196.191.157 & ip route 10.201.20.49 255.255.255.255 10.196.192.93), sip traffic coming from one link can’t be sure to send it back to the same link.
-          SIP media & signaling control binding issue, as CUBE support sip binding using one interface only “one IP Address”, if we not using binding commands on the CUBE we can’t receive any calls though any link.
We have two options:
SP to send both traffic on the same trunk link
Or
Have another CUBE for the second link.
Attached network diagram.
Any solution?????
Regards,
Ahmed Rizk

I didn't mean NAT CUCM, I meant the interface towards it. But since you're using a single interface then yes that is what you NAT. You have a lot going on in that config. Probably a lot more than you need. Like I said you should work on this in two legs. CUBE to ITSP, and then CUCM to CUBE. You're trying to make the whole thing work in one shot which is going to cause you some headaches.
Install XLite free version. In the account settings set your UserID to a generic 10 digit phone number, domain to something generic, then at the bottom set the Proxy Address to the IP of your CUBE. The media ports will be negotiated dynamically between the CUBE and the ITSP. Since you said you're not registering you will also need to give the ITSP YOUR peer IP (this is how they secure the trunk) which is whatever IP you're sourcing from when you leave your network (what you're NAT'ing the CUBE to).
For testing, reduce your config to something like this:
voice service voip
allow-connections sip to sip
allow-connections h323 to sip
no supplementary-service sip moved-temporarily
no supplementary-service sip refer
signaling forward none
sip
dial-peer voice 10 voip
description CUBE_TO_ITSP
session protocol sipv2
session target ipv4:SIGNALING IP PROVIDED BY ITSP
destination-pattern [2-9].........
codec g711ulaw
dtmf-relay rtp-nte sip-notify
no vad
dial-peer voice 20 voip
description ITSP_TO_CUBE
destination-pattern .
session protocol sipv2
session target ipv4:Eventually your CUCM IP...for now set it to your computers IP.
codec g711ulaw
dtmf-relay rtp-nte sip-notify
no vad
Use XLite to place a phone call from your PC (if you have a mic and speakers you can have audio if the call connects). This should come pretty close to getting your outward leg established. Once you get this part working you can add in more codecs and translation profiles if you want. Let me know what happens. Include any debug or packet cap results if you can.

Connecting two computers using ethernet

How do i connect two computers to transfer my iphoto library using ethernet? How do I know how long it will run?

Why not connect the two using Firewire Target mode. Apple explains how in:
http://support.apple.com/kb/HT1661

I want to connect two pair for earphones to my IPAD at the same time is there a recommended splitter or amplifier I should use?

I want to connect two pair for earphones to my IPAD at the
same time is there a recommended splitter or amplifier I should use?

tracer took the words out of my mouth. Doing the search I found this one amongst many others.
http://store.apple.com/us/product/HA828ZM/A/belkin-audio-splitter-35mm-m2x35mm-f

Is it possible to use the same Switch for two different clusters.

I have 10g Rac setup on windows.
Now I am planning to install 11gR2 on different servers.
Is it possible to use the same Switch for two different clusters.

user9198027 wrote:
I have 10g Rac setup on windows.
Now I am planning to install 11gR2 on different servers.
Is it possible to use the same Switch for two different clusters.
Yes. Technically there will not be any conflict as long as the private addresses used by the 2 clusters do not collide, and provided that the switch's port capacity and bandwidth will not be exceeded.
Your NA (netadmin) can also configure the switch to separate the 2 Interconnects from one another (called partitioning when using Infiniband) - if the switch supports such features.
A major consideration is not to make the switch, public. That typically cause a range of problems and can have a serious impact on an Interconnect. But using 2 private networks on the same infrastructure should not have the same problems - if configured and implemented correctly.

Connecting two Coherence Clusters using Replication

Similar Messages

Maybe you are looking for