Re: Problems With Factory Configurations In Karaf 3.0.1
Hello John, It looks like it is fixed on the 3.0.x branch: https://github.com/apache/karaf/blob/karaf-3.0.x/features/core/src/main/java/org/apache/karaf/features/internal/FeatureConfigInstaller.java Guillaume reverted the offending change on July 11: https://github.com/apache/karaf/commit/c31edca04692d8127989acbcdadc8157d1390f2a The same revert was done on master but Achim appears to have remade the change here: https://github.com/apache/karaf/commit/21089127c89ddae275d495607c422dda8ffec474 Achim - was that intentional? I am not sure if David was referring to me in some way when he said there is a bunch of misinformation in some of the comments on this thread. I don't really care about the philosophical reasons for one particular factory pid format or the other. I just know with this change my ManagedServiceFactories and Declarative Service components could not find their factory configurations anymore...which thus made it impossible for me at least to upgrade to Karaf 3. rant Karaf is a great project but the surprise regressions and the long variable delays between versions (which include fixes for these regressions) is very frustrating. What makes it worse for me is that in the version I was using (Karaf 2.3.2) editing factory configurations (not stored in etc) from the Karaf command line console is broken causing me hours upon hours of misery as the workaround for it (copy the factory configuration to a file in the etc directory, edit the file, delete the old factory configuration, restart karaf) is very error prone (I can't ask people to use the web console because I need to use Pax Web 3 with Karaf 2.3.2). I keep being told how karaf is too hard to use. Oh well... /rant thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4035683.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Problems With Factory Configurations In Karaf 3.0.1
OK, added:
https://issues.apache.org/jira/browse/KARAF-3100
and:
https://github.com/apache/karaf/pull/43
to allow karaf to write feature configurations to file in the ${karaf.etc}
dir. Let me know if this makes sense.
thanks,
Gareth
--
View this message in context:
http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4034002.html
Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Problems With Factory Configurations In Karaf 3.0.1
Just to clarify here. I was considering implementing this feature (it doesn't look that hard). Just wondering whether it is something that would be accepted if the implementation is OK. thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4033983.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Problems With Factory Configurations In Karaf 3.0.1
Hello Achim, So I can assume that what I am seeing is not the expected behaviour then? If yes, I will open a JIRA. Note that it messes up configuration for other components as well. e.g. This is what I see when I install hawtio-git which can't be correct: Pid:git.2bb62379-f85c-4846-bc6e-557ac591738b FactoryPid: git BundleLocation: null Properties: hawtio.config.dir = ./etc org.apache.karaf.features.configKey = hawtio-git service.factoryPid = git service.pid = git.2bb62379-f85c-4846-bc6e-557ac591738b thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4033944.html Sent from the Karaf - User mailing list archive at Nabble.com.
RE: Problems With Factory Configurations In Karaf 3.0.1
Hello Benjamin,
Thanks for doing the testing! I think I can see the relevant commit:
https://github.com/apache/karaf/commit/b398c7690a09b02ad0f900fa4d6f51308f5970aa
features/core/src/main/java/org/apache/karaf/features/internal/FeaturesServiceImpl.java
protected Configuration createConfiguration(ConfigurationAdmin
configurationAdmin,
String pid, String
factoryPid) throws IOException, InvalidSyntaxException {
if (factoryPid != null) {
-return configurationAdmin.createFactoryConfiguration(pid,
null);
+return
configurationAdmin.createFactoryConfiguration(factoryPid, null);
} else {
return configurationAdmin.getConfiguration(pid, null);
}
If I understand the code in this file correctly, if I put this in my feature
file:
config name=foo.bar-unencrypted
the pid variable here is foo.bar and the factoryPid variable is
unencrypted. Perhaps this change should be reverted?
thanks,
Gareth
--
View this message in context:
http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4033949.html
Sent from the Karaf - User mailing list archive at Nabble.com.
RE: Problems With Factory Configurations In Karaf 3.0.1
Hello Benjamin,
I think theproblem is that the names of the variables are
incorrect/confusing. Look at the parsePid method in the same file
private String[] parsePid(String pid) {
int n = pid.indexOf('-');
if (n 0) {
String factoryPid = pid.substring(n + 1);
pid = pid.substring(0, n);
return new String[]{pid, factoryPid};
} else {
return new String[]{pid, null};
}
}
The factory pid is definitely not the string after the - as suggested
here.
Gareth
Gareth
--
View this message in context:
http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4033954.html
Sent from the Karaf - User mailing list archive at Nabble.com.
RE: Problems With Factory Configurations In Karaf 3.0.1
This parse pid really needs to return three variables (possibly):
private String[] parsePid(String pid) {
int n = pid.indexOf('-');
if (n 0) {
String instance = pid.substring(n + 1);
String factoryPid = pid.substring(0, n);
return new String[]{null, factoryPid, instance}; // pid is null
as it hasn't been generated yet
} else {
return new String[]{pid, null, null};
}
}
Of course if this change was made additional changes would need to be made
to the file to match this.
thanks,
Gareth
--
View this message in context:
http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4033956.html
Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Problems With Factory Configurations In Karaf 3.0.1
Hello JB, I am missing something here. In Karaf 2.3.2 when I added a configuration like this: config name=foo.bar-unencrypted this created a configuration like this: Pid:foo.bar.704f93e3-56f4-41bc-92d8-c6acf2ae167a FactoryPid: foo.bar BundleLocation: mvn:com.mycompany/foo.bar/1.0.0-SNAPSHOT Properties: host = 0.0.0.0 port = 8881 org.apache.karaf.features.configKey = foo.bar-unencrypted service.factoryPid = foo.bar service.pid = foo.bar.0cfabec4-15f4-4d33-bfa3-c9b4bc0dce66 suggesting that the factory pid should be the string before the - (if there is a -). Was this incorrect? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4033957.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Problems With Factory Configurations In Karaf 3.0.1
Hi Jean-Baptiste, A question (sort of off-topic) - where do the variables in the configuration get replaced? Before this? By config admin? I ask because I am wondering how hard it would be to perhaps redirect this configuration to file (say via a configuration item in org.apache.karaf.features.cfg - perhaps called redirectFeatureConfigsToFile), would you accept it? Could it possibly make it for 3.0.2? This would solve so many problems for me (it allows me to keep all default configuration in the feature file...yet make the configuration easy to edit via configuration file in the etc dir). thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4033959.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Anyone Seen This Declarative Services Error (Karaf 3)?
OK, I didn't see this: https://issues.apache.org/jira/browse/KARAF-2950 Looks like I need to wait for Karaf 3.0.2. thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Anyone-Seen-This-Declarative-Services-Error-Karaf-3-tp4033891p4033913.html Sent from the Karaf - User mailing list archive at Nabble.com.
Problems With Factory Configurations In Karaf 3.0.1
Hello, In my feature file I setup factory configurations like this: feature name=feature-name description=My Wonderful Feature version=1.0.0-SNAPSHOT resolver=(obr) . . config name=foo.bar-unencrypted host=0.0.0.0 port=8081 /config config name=foo.bar-encrypted host=0.0.0.0 port=8881 /config /feature When I install this feature in Karaf 3.0.1, the configuration is appearing like this. I was expecting the factory pid to be foo.bar (as it was for Karaf 2.3.x): Pid:unencrypted.0ed76d9e-5a12-4b2d-bc7f-c02e6bce4b71 FactoryPid: unencrypted BundleLocation: null Properties: org.apache.karaf.features.configKey = foo.bar-unencrypted service.factoryPid = unencrypted service.pid = unencrypted.0ed76d9e-5a12-4b2d-bc7f-c02e6bce4b71 host = 0.0.0.0 port = 8081 Pid:encrypted.6b6f5e66-5fc2-479e-98f1-3873fe16be0d FactoryPid: encrypted BundleLocation: null Properties: org.apache.karaf.features.configKey = foo.bar-encrypted service.factoryPid = encrypted service.pid = encrypted.6b6f5e66-5fc2-479e-98f1-3873fe16be0d host = 0.0.0.0 port = 8881 So my question is - Am I supposed to specify factory configuration in a different way now for Karaf 3? If someone could let me know, it would be a big help. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Problems With Factory Configurations In Karaf 3.0.1
Hello Achim, Thanks for the quick response! I am confused. I understood that if I created a file foo.bar-encrypted.cfg in the etc directory a configuration would be created for factory pid foo.bar. Similarly if I created a configuration like this in a feature file: config name=foo.bar-encrypted . . /config I thought a configuration for factory pid foo.bar would also be created. Have I misunderstood something here? This is the behaviour I saw in karaf 2.3...which doesn't seem to be the case now (at least for adding the configuration for the feature files). Now it is thinking the part after the - is the factory pid. thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Problems-With-Factory-Configurations-In-Karaf-3-0-1-tp4033921p4033928.html Sent from the Karaf - User mailing list archive at Nabble.com.
Where Has The Blueprint Status Gone In Karaf 3.x?
Hello, I am sure I am missing something obvious here but...where has the blueprint status gone? I know there was a separate column in karaf 2.3: karaf@root list -t 0 -s START LEVEL 100 , List Threshold: 0 ID State Blueprint Level Symbolic name . . [ 14] [Active ] [Created ] [ 25] org.apache.karaf.shell.console (2.3.2) [ 15] [Active ] [Created ] [ 28] org.apache.karaf.deployer.spring (2.3.2) [ 16] [Active ] [Created ] [ 28] org.apache.karaf.deployer.blueprint (2.3.2) [ 17] [Active ] [Created ] [ 30] org.apache.karaf.jaas.command (2.3.2) [ 18] [Active ] [Created ] [ 30] org.apache.karaf.jaas.modules (2.3.2) [ 19] [Active ] [Created ] [ 30] org.apache.karaf.shell.dev (2.3.2) How can I see the same information in karaf 3?: 23 | Active | 24 | 3.0.1 | org.apache.karaf.deployer.spring 24 | Active | 24 | 3.0.1 | org.apache.karaf.deployer.blueprint 40 | Active | 30 | 3.0.1 | org.apache.karaf.shell.console 41 | Active | 30 | 3.0.1 | org.apache.karaf.jaas.modules 58 | Active | 30 | 3.0.1 | org.apache.karaf.jaas.command thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Where-Has-The-Blueprint-Status-Gone-In-Karaf-3-x-tp4033890.html Sent from the Karaf - User mailing list archive at Nabble.com.
Anyone Seen This Declarative Services Error (Karaf 3)?
Hello, I am installing the following base features when starting karaf 3.0.1: # # Comma separated list of features to install at startup # featuresBoot=config,standard,region,package,kar,obr,ssh,management,scr,war # # Defines if the boot features are started in asynchronous mode (in a dedicated thread) # featuresBootAsynchronous=false After first start, I seem to be getting a weird error when trying to run scr:list. I am seeing the following in the log: 2014-06-30 22:41:29,330 | DEBUG | lixDispatchQueue | o.a.f.scr | []:[]:[]:[] | 67 - org.apache.felix.scr - 1.8.2 | FrameworkEvent ERROR - org.apache.felix.scr org.osgi.framework.ServiceException: Service cannot be cast: org.apache.felix.scr.impl.ScrGogoCommand at org.apache.felix.framework.ServiceRegistrationImpl.getFactoryUnchecked(ServiceRegistrationImpl.java:332) ~[na:na] at org.apache.felix.framework.ServiceRegistrationImpl.getService(ServiceRegistrationImpl.java:219) ~[na:na] at org.apache.felix.framework.ServiceRegistry.getService(ServiceRegistry.java:320) ~[na:na] at org.apache.felix.framework.Felix.getService(Felix.java:3568) ~[na:na] at org.apache.felix.framework.BundleContextImpl.getService(BundleContextImpl.java:468) ~[na:na] at org.apache.karaf.shell.console.completer.CommandsCompleter.unProxy(CommandsCompleter.java:298) ~[na:na] at org.apache.karaf.shell.console.completer.CommandsCompleter.checkData(CommandsCompleter.java:234) ~[na:na] at org.apache.karaf.shell.console.completer.CommandsCompleter.complete(CommandsCompleter.java:86) ~[na:na] at org.apache.karaf.shell.console.impl.jline.CompleterAsCompletor.complete(CompleterAsCompletor.java:32) ~[na:na] at jline.console.ConsoleReader.complete(ConsoleReader.java:3077) ~[na:na] at jline.console.ConsoleReader.readLine(ConsoleReader.java:2501) ~[na:na] at jline.console.ConsoleReader.readLine(ConsoleReader.java:2162) ~[na:na] at org.apache.karaf.shell.console.impl.jline.ConsoleImpl.readAndParseCommand(ConsoleImpl.java:280) ~[na:na] at org.apache.karaf.shell.console.impl.jline.ConsoleImpl.run(ConsoleImpl.java:207) ~[na:na] at java.lang.Thread.run(Thread.java:744) ~[na:1.7.0_51] at org.apache.karaf.shell.console.impl.jline.ConsoleFactoryService$3.doRun(ConsoleFactoryService.java:126) ~[na:na] at org.apache.karaf.shell.console.impl.jline.ConsoleFactoryService$3$1.run(ConsoleFactoryService.java:117) ~[na:na] at java.security.AccessController.doPrivileged(Native Method) ~[na:1.7.0_51] at org.apache.karaf.jaas.modules.JaasHelper.doAs(JaasHelper.java:47) ~[na:na] at org.apache.karaf.shell.console.impl.jline.ConsoleFactoryService$3.run(ConsoleFactoryService.java:115) ~[na:na] Anyone seen an error like this before? Just trying to figure out what I may have done wrong here. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Anyone-Seen-This-Declarative-Services-Error-Karaf-3-tp4033891.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Require-Capabilties and Karaf 2.3
Hello Achim, OK, I understand my problem. It is because I am using equinox. Christian mentioned the issue here: http://www.mail-archive.com/users@felix.apache.org/msg15464.html And I understand Christian opened an issue about it (with information about a workaround as well - I haven't tried it yet): https://issues.apache.org/jira/browse/KARAF-3069 Thanks for your help! Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Require-Capabilties-and-Karaf-2-3-tp4033665p4033859.html Sent from the Karaf - User mailing list archive at Nabble.com.
Registering My Own Connection Factory For The Karaf JMS Service
Hello, I currently register my own JMS connection factories as services. It would be nice if I could take advantage of the new JMS command line tools in karaf (especially since the activemq web console has had problems for a long time with equinox - ugh). I started playing around with the karaf jms tools and I see these service parameters registered: karaf@root() ls javax.jms.ConnectionFactory [javax.jms.ConnectionFactory] - osgi.service.blueprint.compname = pooledConnectionFactory name = jms-default osgi.jndi.service.name = jms/jms-default service.id = 1167 Provided by : Bundle 168 To get my connection factory seen by the karaf jms tools, I assume I just need to make sure that I have set osgi.jndi.service.name to jms/my connection factory name in my service properties? Do these tools explicitly require a pooled/cached connection factory...or it doesn't matter? thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Registering-My-Own-Connection-Factory-For-The-Karaf-JMS-Service-tp4033861.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Require-Capabilties and Karaf 2.3
Hello Achim,
Is this the configuration? I see this in both the 2.3.3 and 2.3.5 files.
thanks,
Gareth
org.osgi.framework.system.capabilities= \
${eecap-${java.specification.version}}, \
service-reference;effective:=active;objectClass=org.osgi.service.packageadmin.PackageAdmin,
\
service-reference;effective:=active;objectClass=org.osgi.service.startlevel.StartLevel,
\
service-reference;effective:=active;objectClass=org.osgi.service.url.URLHandlers,
\
${services-${karaf.framework}}
eecap-1.7= osgi.ee; osgi.ee=OSGi/Minimum;
version:ListVersion=1.0,1.1,1.2, \
osgi.ee; osgi.ee=JavaSE;
version:ListVersion=1.0,1.1,1.2,1.3,1.4,1.5,1.6,1.7
eecap-1.6= osgi.ee; osgi.ee=OSGi/Minimum;
version:ListVersion=1.0,1.1,1.2, \
osgi.ee; osgi.ee=JavaSE;
version:ListVersion=1.0,1.1,1.2,1.3,1.4,1.5,1.6
eecap-1.5= osgi.ee; osgi.ee=OSGi/Minimum;
version:ListVersion=1.0,1.1,1.2, \
osgi.ee; osgi.ee=JavaSE; version:ListVersion=1.0,1.1,1.2,1.3,1.4,1.5
eecap-1.4= osgi.ee; osgi.ee=OSGi/Minimum;
version:ListVersion=1.0,1.1,1.2, \
osgi.ee; osgi.ee=JavaSE; version:ListVersion=1.0,1.1,1.2,1.3,1.4
eecap-1.3= osgi.ee; osgi.ee=OSGi/Minimum; version:ListVersion=1.0,1.1,
\
osgi.ee; osgi.ee=JavaSE; version:ListVersion=1.0,1.1,1.2,1.3
eecap-1.2= osgi.ee; osgi.ee=OSGi/Minimum; version:ListVersion=1.0,1.1,
\
osgi.ee; osgi.ee=JavaSE; version:ListVersion=1.0,1.1,1.2
--
View this message in context:
http://karaf.922171.n3.nabble.com/Require-Capabilties-and-Karaf-2-3-tp4033665p4033677.html
Sent from the Karaf - User mailing list archive at Nabble.com.
Require-Capabilties and Karaf 2.3
Hello, I recently updated bndtools 2.3. in my build environment and it appears to have started adding the following to all my bundles: Require-Capability: osgi.ee;filter:=((osgi.ee=JavaSE)(version=1.7)) When I try and install in karaf 2.3.3 I am getting the following exception: Error executing command: Could not start bundle my bundle: The bundle my bundle name [138] could not be resolved. Reason: Missing Constraint: Require-Capability: osgi.ee; filter=((osgi.ee=JavaSE)(version=1.7)) Is this Require-Capability supported in karaf 2.3.x? If yes, any commands in 2.3 which would allow me to check these capabilities. Just wondering whether I should find a way to remove the Require-Capability from my bundle temporarilty (I plan to upgrade to karaf 3.x after 3.0.2 comes out). thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Require-Capabilties-and-Karaf-2-3-tp4033665.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: WABs and Web-ContextPath
Hello Scott, There are other ways to do this in OSGi/Karaf. But to use these other methods you need to use the http service or the pax web implementation of the whiteboard pattern to register servlet contexts/servlets/filters etc. One way to implement your requirement would be to create one bundle to manage the registering of resources with pax web. Each Servlet implementation bundle could have a special manifest flag/special file which identifies it as part of the application. The main bundle would then be a bundle listener which looks for this special file/manifest flag and when it finds it it looks inside the bundle and registers relevant resources with pax web under the main servlet context. I do this already and it works well for me. Let me know if this answers your question. regards, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/WABs-and-Web-ContextPath-tp4032881p4032948.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Karaf 3.0.0 Questions
Hello Jean-Baptiste, You are right. I was getting the URL wrong :). I only brought this up because I thought the new message was confusing. The karaf 2.3.3 message was clear to me: karaf@root features:addurl mvn:org.apache.activemq/activemq-karaf/5.9.0/features/xml Error executing command: Unable to add repositories: URL [mvn:org.apache.activemq/activemq-karaf/5.9.0/features/xml] could not be resolved. The 3.0.0 error Could not find artifact xxx in defaultlocal (file:/home/user/.m2/repository) suggests that I don't have any remote repositories configured. I will take a look at the pax url code. thanks again, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Karaf-3-0-0-Questions-tp4030800p4030814.html Sent from the Karaf - User mailing list archive at Nabble.com.
Karaf 3.0.0 Questions
Hello, I would first like to say thankyou for releasing Karaf 3.0.0. Just in time for Christmas! Great work! Of course, as soon as you released it I decided to download and play with it. After my initial play, I have a few questions: (1) The first thing I tried to do was install activemq, and of course I got the URL wrong the first time (I also know now that I don't need to know the full URL for activemq as it is in the features.repo file). I mention this because the error message appeared confusing: karaf@root() feature:repo-add mvn:org.apache.activemq/activemq-karaf/5.9.0/features/xml Adding feature url mvn:org.apache.activemq/activemq-karaf/5.9.0/features/xml Error executing command: Error resolving artifact org.apache.activemq:activemq-karaf:features:xml:5.9.0: Could not find artifact org.apache.activemq:activemq-karaf:features:xml:5.9.0 in defaultlocal (file:/home/gcollins/.m2/repository/) Is this error message coming from karaf or pax-url? When I saw this the first time I thought it meant that external repositories (i.e maven central) were not configured. Is it confusing to anyone else? (2) I understand that activemq uses hawtio now for the webconsole so I tried to install that: karaf@root() feature:install hawtio Error executing command: Could not start bundle mvn:io.hawt/hawtio-karaf-terminal/1.2.1/war in feature(s) hawtio-karaf-terminal-1.2.1: The bundle io.hawt.hawtio-karaf-terminal_1.2.1 [164] could not be resolved. Reason: Missing Constraint: Import-Package: org.apache.karaf.shell.console.jline; version=[2.2.0,3.0.0) Is this a known hawtio issue? If not, where should I open an issue for this? (3) Any plans to update to javax.annotations 1.2? I ask because this is a dependency for jersey 2 (I am mulling upgrading from jetty 1.x). Or is this something irrelevant to karaf (i.e. this is purely an app level responsibility)? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Karaf-3-0-0-Questions-tp4030800.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Updating Managed Service Factory Config Via config: commands?
Issue created (better late than never): https://issues.apache.org/jira/browse/KARAF-2366 thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Updating-Managed-Service-Factory-Config-Via-config-commands-tp4028322p4029090.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Uninstalling Declarative Services Bundles Which Require Configuration
OK, I assume it is a defect then. JIRA added: https://issues.apache.org/jira/browse/KARAF-2367 thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Uninstalling-Declarative-Services-Bundles-Which-Require-Configuration-tp4029022p4029091.html Sent from the Karaf - User mailing list archive at Nabble.com.
Uninstalling Declarative Services Bundles Which Require Configuration
Hello, I am seeing behaviour in karaf which I did not expect. I uninstalled a bundle which used declarative services...and reinstalled it from a different location (old location - maven repository in nexus, new location - local file system). When I ran config:list for the pid after doing this, it still showed that the config was bound to the old bundle, even though it was uninstalled (which of course meant the new bundle could not be activated using declarative services). This behaviour persisted even after a restart of karaf. To fix I had to remove and re-add the configuration. I shouldn't need to do this, should I? This is a bug? I am using karaf 2.3.1 with equinox. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Uninstalling-Declarative-Services-Bundles-Which-Require-Configuration-tp4029022.html Sent from the Karaf - User mailing list archive at Nabble.com.
Updating Managed Service Factory Config Via config: commands?
Hi, I am trying to update factory config via config: commands (Karaf 2.3.1). This config was created by an entry in a feature file thus a backing config file does not exist. i.e.: Pid:com.mycompany.myservice.327d4cd4-4704-4190-9e51-26f5f0b91435 FactoryPid: com.mycompany.myservice BundleLocation: mvn:com.mycompany/mycompany/1.0.0-SNAPSHOT Properties: myservice.host = 0.0.0.0 myservice.port = 8080 org.apache.karaf.features.configKey = com.mycompany.myservice-healthcheck service.factoryPid = com.mycompany.myservice service.pid = com.mycompany.myservice.327d4cd4-4704-4190-9e51-26f5f0b91435 If I try and edit this config. e.g.: config:edit com.mycompany.myservice-healthcheck config:propset myservice.port = 8081 config:update I get a new config: Pid:com.mycompany.myservice.8cbfea87-f66e-44b4-b94a-11fdb14c235f FactoryPid: com.mycompany.myservice BundleLocation: mvn:com.antennasoftware/gravity.clientAP.blueprint/1.0.0-SNAPSHOT Properties: service.pid = com.mycompany.myservice.8cbfea87-f66e-44b4-b94a-11fdb14c235f myservice.port = 8081 service.factoryPid = com.mycompany.myservice felix.fileinstall.filename = file:/home/karaf/apache-karaf-2.3.1/etc/com.mycompany.myservice-healthcheck.cfg If I go and specify the pid exactly: config:edit com.mycompany.myservice.327d4cd4-4704-4190-9e51-26f5f0b91435 config:propset myservice.port 8081 config:update Nothing happens. Am I missing something here? BTW, if the factory configuration is backed by a file everything works fine. And if I update a non-factory not backed by a file, a file gets created after the update. If someone could let me know what I am doing wrong, it would be much appreciated. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Updating-Managed-Service-Factory-Config-Via-config-commands-tp4028322.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Updating Managed Service Factory Config Via config: commands?
Hi Jean-Baptiste, Achim, Thanks very much for the responses! When I look at the help I see: OPTIONS -f, --use-file Configuration lookup using the filename instead of the pid which suggests this has something to do with the file already being there. Does this need an update? Anyway I tried to run this command. Result below: karaf@root config:edit -f com.mycompany.myservice-healthcheck Error executing command: java.lang.NullPointerException 2013-03-26 21:10:08,231 | INFO | []:[] | Thread-8220 | Console | araf.shell.console.jline.Console 198 | 14 - org.apache.karaf.shell.console - 2.3.1 | Exception caught while executing command java.lang.NullPointerException at org.apache.karaf.shell.config.ConfigCommandSupport.findConfigurationByFileName(ConfigCommandSupport.java:115) at org.apache.karaf.shell.config.EditCommand.doExecute(EditCommand.java:50) at org.apache.karaf.shell.config.ConfigCommandSupport.doExecute(ConfigCommandSupport.java:68) at org.apache.karaf.shell.console.OsgiCommandSupport.execute(OsgiCommandSupport.java:38) at org.apache.felix.gogo.commands.basic.AbstractCommand.execute(AbstractCommand.java:35) at org.apache.felix.gogo.runtime.CommandProxy.execute(CommandProxy.java:78) at org.apache.felix.gogo.runtime.Closure.executeCmd(Closure.java:474) at org.apache.felix.gogo.runtime.Closure.executeStatement(Closure.java:400) at org.apache.felix.gogo.runtime.Pipe.run(Pipe.java:108) at org.apache.felix.gogo.runtime.Closure.execute(Closure.java:183) at org.apache.felix.gogo.runtime.Closure.execute(Closure.java:120) at org.apache.felix.gogo.runtime.CommandSessionImpl.execute(CommandSessionImpl.java:89) at org.apache.karaf.shell.console.jline.Console.run(Console.java:174) at java.lang.Thread.run(Unknown Source)[:1.7.0_05] at org.apache.karaf.shell.ssh.ShellFactoryImpl$ShellImpl$4.doRun(ShellFactoryImpl.java:144)[28:org.apache.karaf.shell.ssh:2.3.1] at org.apache.karaf.shell.ssh.ShellFactoryImpl$ShellImpl$4$1.run(ShellFactoryImpl.java:135) at java.security.AccessController.doPrivileged(Native Method)[:1.7.0_05] at javax.security.auth.Subject.doAs(Unknown Source)[:1.7.0_05] at org.apache.karaf.shell.ssh.ShellFactoryImpl$ShellImpl$4.run(ShellFactoryImpl.java:133)[28:org.apache.karaf.shell.ssh:2.3.1] thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Updating-Managed-Service-Factory-Config-Via-config-commands-tp4028322p4028325.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Logback and Apache Karaf
Isn't it though a separate bundle (pax-logging-service vs. pax-logging-logback)? From looking at the code it looks like whilst it uses the same pid as the pax-logging-service, pax-logging-logback works a little differently (the org.ops4j.pax.logging configuration contains a pointer to the real logback configuration). I can do all this myself (download pax-logging-logback to the embedded repository, change startup properties to use pax-logging-logback instead of pax-logging-service, change the default org.ops4j.pax.logging.properties file and add a logback.xml). It would be nice though if karaf could make it easier (like it does with the equinox/felix switch). I am actually surprised that no-one else asked for this. Is logback not used much? thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Logback-and-Apache-Karaf-tp4028183p4028195.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Logback and Apache Karaf
Hi Ryan, I am curious - How did you install it? thanks, Gareth On Thu, Mar 14, 2013 at 6:26 PM, Ryan Moquin [via Karaf] ml-node+s922171n4028206...@n3.nabble.com wrote: FYI, I tried the pax-logging-logback library (not sure if it was 1.7.0 or 1.7.1 or something), but it went crazy when I installed it. Literally just about every bundle version of logback that was ever made ended up installed and activated there were some other weird things. I ended up deciding just to stick with what's there by default though I would like to use logback somehow. On Thu, Mar 14, 2013 at 11:16 AM, Guillaume Nodet [hidden email] wrote: Another drawback of switching to logback would be that the log:set command which allow changing the logger levels would not work anymore. On Thu, Mar 14, 2013 at 9:03 AM, Achim Nierbeck [hidden email] wrote: No not really, cause right now we rather use the ConfigurationAdmin Service way of configuring the logger. But I think you're able to configure it in the org.ops4j.pax.logging.cfg file to use a logback config. regards, Achim 2013/3/14 Gareth [hidden email] Hi, Any plans to make it easy to switch the Pax Logging (now 1.7.0) now bundled with Karaf to use Logback instead? e.g. like having a: karaf.loggingbackend = log4j/logback I ask because I was interested in using logback specific features (e.g. log4j 1.2 does not support Markers). thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Logback-and-Apache-Karaf-tp4028183.html Sent from the Karaf - User mailing list archive at Nabble.com. -- Apache Karaf http://karaf.apache.org/ Committer PMC OPS4J Pax Web http://wiki.ops4j.org/display/paxweb/Pax+Web/ Committer Project Lead OPS4J Pax for Vaadin http://team.ops4j.org/wiki/display/PAXVAADIN/Home Commiter Project Lead blog http://notizblog.nierbeck.de/ -- Guillaume Nodet Red Hat, Open Source Integration Email: [hidden email] Web: http://fusesource.com Blog: http://gnodet.blogspot.com/ If you reply to this email, your message will be added to the discussion below: http://karaf.922171.n3.nabble.com/Logback-and-Apache-Karaf-tp4028183p4028206.html To unsubscribe from Logback and Apache Karaf, click here. NAML -- View this message in context: http://karaf.922171.n3.nabble.com/Logback-and-Apache-Karaf-tp4028183p4028207.html Sent from the Karaf - User mailing list archive at Nabble.com.
Logback and Apache Karaf
Hi, Any plans to make it easy to switch the Pax Logging (now 1.7.0) now bundled with Karaf to use Logback instead? e.g. like having a: karaf.loggingbackend = log4j/logback I ask because I was interested in using logback specific features (e.g. log4j 1.2 does not support Markers). thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Logback-and-Apache-Karaf-tp4028183.html Sent from the Karaf - User mailing list archive at Nabble.com.
Why Does Spring JMS Depend On Spring Web?
Hi, A quick question - in looking at the Karaf feature files, I see that spring-jms depends on spring-web being installed. Any reason why this is the case? I don't see the connection. In a dev environment (in Eclipse, outside Karaf) I run quite happily without installing the Spring Web bundles. Just trying to understand. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Why-Does-Spring-JMS-Depend-On-Spring-Web-tp4028018.html Sent from the Karaf - User mailing list archive at Nabble.com.
Karaf 3.x Plans?
Hi, Are there plans to release Karaf 3.0 in the near future? I ask because I am interested in using a more current Pax Web release with Karaf (the Pax Web 1.1/Jetty 7 release is now a little old). I am curious - is Karaf 3.x just waiting for new releases of dependencies (e.g. Aries, XBean?)...or are there major features that still need to be implemented? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Karaf-3-x-Plans-tp4027587.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Gogo vs. Karaf Commands
Hi Gokturk, I am interested in seeing this happen as well. Perhaps you can share what you did so far on github? Just make a fork of apache karaf then push your changes to your fork? thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Gogo-vs-Karaf-Commands-tp4027219p4027303.html Sent from the Karaf - User mailing list archive at Nabble.com.
Gogo vs. Karaf Commands
Hello, I am curious - What is the difference between a Gogo command and a Karaf command? I ask because I wrote for myself a couple of gogo commands. If I run vanilla OSGi (e.g. equinox) with gogo they run fine. The commands also show up if I run gogo help. If I run these same gogo commands in karaf, they also appear to work. However, the commands don't show up in Karaf help and I cannot take advantage of other advanced karaf console features like autocomplete with these commands. Any reason why this would not work? Is there perhaps just an additional service property I can add which will allow me to take advantage of the additional Karaf console support? Having looked at a few Karaf console command examples, I see they all extend the Karaf class OsgiCommandSupport, which I prefer not to do as I would like to avoid having to explicitly depend on Karaf classes. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Gogo-vs-Karaf-Commands-tp4027219.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Questions About the 2.3.0 Web Console
Hello Jean-Baptiste, Thanks very much for the quick response! My apologies about mentioning the shutdown issue. I should have checked JIRA. I thought the web console problem may be a timing issue. However I tried stopping and starting the web console bundle, uninstalling/reinstalling the web console feature but I still haven't been able to get the web console plugins to work. I don't see anything very exciting in the logs. Anything else useful I can try? thanks again, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Questions-About-the-2-3-0-Web-Console-tp4026419p4026438.html Sent from the Karaf - User mailing list archive at Nabble.com.
Questions About the 2.3.0 Web Console
Hi, Thanks for all the hard work in releasing Karaf 2.3.0! Anyway, I am just starting to try out the new release and I am seeing a couple of quirks: (1) It appears to hang on shutdown. Not a big issue as I can always use ^C. (2) The webconsole appears to have a few problems: (a) The following pages come up blank - Admin, Features and Gogo. Do I need to do anything special to see these pages? (b) I am not seeing the Components plugin in the console. Do I need to do something special here? In the past I just needed to make sure that Felix Declarative Services is started before the webconsole. I am seeing this behaviour running on CentOS 6.2, JDK 1.7.05 using equinox and the following boot feature list: # # Comma separated list of features to install at startup # featuresBoot=config,ssh,management,scr,my own feature to load pax url classpath,obr,kar,war,webconsole thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Questions-About-the-2-3-0-Web-Console-tp4026419.html Sent from the Karaf - User mailing list archive at Nabble.com.
Can Fixes For These Issues Get Into The Next Release?
Hello, Can fixes for the following issues get into the next release (whether 2.2.10 or 2.3.0 or both)?: https://issues.apache.org/jira/browse/KARAF-1796 https://issues.apache.org/jira/browse/KARAF-1759 https://issues.apache.org/jira/browse/KARAF-1765 I have added suggested fixes for all these defects based on the 2.2.x stream (the fixes work for me). If there is something additional I can do to help make sure these fixes get in please let me know. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Can-Fixes-For-These-Issues-Get-Into-The-Next-Release-tp4026234.html Sent from the Karaf - User mailing list archive at Nabble.com.
More Questions (Mostly On Features)
Hello, A few questions (this will be a long email): (1) I see in the documentation that you should be able to do this to do this in a feature file: config name=com.foo.bar-alias myProperty = myValue /config where com.foo.bar is a factory pid. What version of karaf should this be available for? Should this work in 2.2.x (normal pids work - I don't see anything for these factory pids)? I see something in 2.3.x (I see something in the UI but nothing goes to the etc directory)...though I haven't got it to work yet (my bundle doesn't see it). Is there any requirement to have a metatype.xml for this to get it to work? (2) Should cellar 2.2.4 work with karaf 2.3.x? I am trying it out as I would like to play with an OSGi 4.3 environment. Would karaf 3.0.x and cellar trunk be better? Anyway, here is the exception I get when I try this: 2012-08-31 00:44:04,986 | ERROR | rint Extender: 1 | BlueprintContainerImpl | container.BlueprintContainerImpl 362 | 9 - org.apache.aries.blueprint.core - 1.0.0 | Unable to start blueprint container for bundle org.apache.karaf.cellar.features org.osgi.service.blueprint.container.ComponentDefinitionException: Unable to intialize bean synchronizer at org.apache.aries.blueprint.container.BeanRecipe.runBeanProcInit(BeanRecipe.java:710)[9:org.apache.aries.blueprint.core:1.0.0] at org.apache.aries.blueprint.container.BeanRecipe.internalCreate2(BeanRecipe.java:820)[9:org.apache.aries.blueprint.core:1.0.0] at org.apache.aries.blueprint.container.BeanRecipe.internalCreate(BeanRecipe.java:783)[9:org.apache.aries.blueprint.core:1.0.0] at org.apache.aries.blueprint.di.AbstractRecipe$1.call(AbstractRecipe.java:79)[9:org.apache.aries.blueprint.core:1.0.0] at java.util.concurrent.FutureTask$Sync.innerRun(Unknown Source)[:1.7.0_05] at java.util.concurrent.FutureTask.run(Unknown Source)[:1.7.0_05] . . Caused by: java.lang.NullPointerException at org.apache.karaf.cellar.features.FeaturesSynchronizer.pull(FeaturesSynchronizer.java:103)[101:org.apache.karaf.cellar.features:2.2.4] at org.apache.karaf.cellar.features.FeaturesSynchronizer.init(FeaturesSynchronizer.java:53)[101:org.apache.karaf.cellar.features:2.2.4] at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)[:1.7.0_05] at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)[:1.7.0_05] at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)[:1.7.0_05] at java.lang.reflect.Method.invoke(Unknown Source)[:1.7.0_05] (3) It appears that features install bundles in the following way: install bundle start bundle install bundle start bundle . . Any reason why you don't do this?: install bundle install bundle install bundle . . start bundle start bundle start bundle The advantage of doing it this way is that you can be sure all optional dependencies/fragments are resolved before starting any bundle (thus you don't have to worry about ordering the bundles). I have a patch...and I will open a JIRA unless someone thinks this is foolish. This solves for me problems with getting fragments resolved (which are started after their bundle hosts) (4) Another idea for a feature for the feature file: Currently you can do this: configfile finalname=/etc/jetty.xmlmvn:org.apache.karaf/apache-karaf/2.2.5/xml/jettyconfig/configfile It would be nice if you could also lay down a tar to a directory as well (I have a whole host of config files to lay down): tarfile directory=/etc/myconfigfoldermvn:com.mycompany/myproject/1.0.0/tar.gz/myconfig/configfile Is this a reasonable idea? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/More-Questions-Mostly-On-Features-tp4025892.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Get environment variable
Thanks Scott! I was interested in this answer as well. This needs to be documented somewhere as it is an important feature. Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Get-environment-variable-tp4025807p4025832.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Features and OBR (was Re: A Basic OBR Question)
Hello Jean-Baptiste, I have added a patch to fix KARAF-1759. I also added a new defect (KARAF-1765) because I believe the shell should make installing optional dependencies optional (rather than always installing). As well the shell should be consistent with the OBR Web UI and the Karaf OBR resolver in not installing optional dependencies by default (I felt it was confusing). A patch has also been added for this issue. Both patches were tested against a karaf 2.2.9 with shell.obr and features.obr updated for the changes. If these patches look reasonable, if you could put them into your next release it would be a big help! I know you just released 2.2.9, but when do you think that might be? I am just kind of keen on the KARAF-1759 fix :). thanks again, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/A-Basic-OBR-Question-tp4025744p4025800.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Features and OBR (was Re: A Basic OBR Question)
OK, now I think I understand. Looking at ObrResolver.java and
BundleInfo.java:
public ListBundleInfo resolve(Feature feature) throws Exception {
.
.
for (BundleInfo bundleInfo : feature.getBundles()) {
.
.
}
.
.
for (Resource res : deploy) { list of resolved resources from OBR
.
.
if (info == null) {
info = new BundleInfoImpl(res.getURI());
}
bundles.add(info);
}
return bundles;
}
If I now go to BundleInfoImpl.java start is implicitly set to false by
default (it should really be set explicitly) which explains why the
dependent bundles are not starting:
public class BundleInfoImpl implements BundleInfo {
private int startLevel;
private String location;
private boolean start;
private boolean dependency;
.
.
.
public BundleInfoImpl(String location) {
this.location = location;
}
Would it be possible to add a default start and start-level options to
the obr resolver so I can get these dependent bundles started? Or could this
somehow be set in the feature file as configuration values for each feature
(e.g. feature...resolver=quot;(obr)quot; resolved-start=quot;truequot;
resolved-start-level=quot;60quot;)? This would make the obr resolver far
more useful.
thanks in advance,
Gareth
--
View this message in context:
http://karaf.922171.n3.nabble.com/A-Basic-OBR-Question-tp4025744p4025782.html
Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Features and OBR (was Re: A Basic OBR Question)
Hello Jean-Baptiste, One more question: I am trying to build Karaf 2.2.x trunk (mvn clean install). It is failing when I get to the Karaf Client. I get the following error. Anything obvious I may have messed up which would cause this error? thanks in advance, Gareth [ERROR] Failed to execute goal org.apache.maven.plugins:maven-shade-plugin:1.7:shade (default) on project org.apache.karaf.client: Error creating shaded jar: Invalid signature file digest for Manifest main attributes - [Help 1] org.apache.maven.lifecycle.LifecycleExecutionException: Failed to execute goal org.apache.maven.plugins:maven-shade-plugin:1.7:shade (default) on project org.apache.karaf.client: Error creating shaded jar: Invalid signature file digest for Manifest main attributes at org.apache.maven.lifecycle.internal.MojoExecutor.execute(MojoExecutor.java:217) at org.apache.maven.lifecycle.internal.MojoExecutor.execute(MojoExecutor.java:153) at org.apache.maven.lifecycle.internal.MojoExecutor.execute(MojoExecutor.java:145) at org.apache.maven.lifecycle.internal.LifecycleModuleBuilder.buildProject(LifecycleModuleBuilder.java:84) at org.apache.maven.lifecycle.internal.LifecycleModuleBuilder.buildProject(LifecycleModuleBuilder.java:59) at org.apache.maven.lifecycle.internal.LifecycleStarter.singleThreadedBuild(LifecycleStarter.java:183) at org.apache.maven.lifecycle.internal.LifecycleStarter.execute(LifecycleStarter.java:161) at org.apache.maven.DefaultMaven.doExecute(DefaultMaven.java:319) at org.apache.maven.DefaultMaven.execute(DefaultMaven.java:156) at org.apache.maven.cli.MavenCli.execute(MavenCli.java:537) at org.apache.maven.cli.MavenCli.doMain(MavenCli.java:196) at org.apache.maven.cli.MavenCli.main(MavenCli.java:141) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) at java.lang.reflect.Method.invoke(Method.java:597) at org.codehaus.plexus.classworlds.launcher.Launcher.launchEnhanced(Launcher.java:290) at org.codehaus.plexus.classworlds.launcher.Launcher.launch(Launcher.java:230) at org.codehaus.plexus.classworlds.launcher.Launcher.mainWithExitCode(Launcher.java:409) at org.codehaus.plexus.classworlds.launcher.Launcher.main(Launcher.java:352) Caused by: org.apache.maven.plugin.MojoExecutionException: Error creating shaded jar: Invalid signature file digest for Manifest main attributes at org.apache.maven.plugins.shade.mojo.ShadeMojo.execute(ShadeMojo.java:553) at org.apache.maven.plugin.DefaultBuildPluginManager.executeMojo(DefaultBuildPluginManager.java:101) at org.apache.maven.lifecycle.internal.MojoExecutor.execute(MojoExecutor.java:209) ... 19 more Caused by: java.lang.SecurityException: Invalid signature file digest for Manifest main attributes at sun.security.util.SignatureFileVerifier.processImpl(SignatureFileVerifier.java:221) at sun.security.util.SignatureFileVerifier.process(SignatureFileVerifier.java:176) at java.util.jar.JarVerifier.processEntry(JarVerifier.java:245) at java.util.jar.JarVerifier.update(JarVerifier.java:199) at java.util.jar.JarFile.initializeVerifier(JarFile.java:323) at java.util.jar.JarFile.getInputStream(JarFile.java:388) at org.apache.maven.plugins.shade.DefaultShader.shade(DefaultShader.java:133) at org.apache.maven.plugins.shade.mojo.ShadeMojo.execute(ShadeMojo.java:494) ... 21 more -- View this message in context: http://karaf.922171.n3.nabble.com/A-Basic-OBR-Question-tp4025744p4025792.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: org.osgi.service.jdbc and Karaf
Is it 5.0 only though? It is in the 4.2 javadoc: http://www.osgi.org/javadoc/r4v42/org/osgi/service/jdbc/DataSourceFactory.html and there is a JDBC Service Specification chapter in the R4.2 enterprise specification (the chapter version was not updated for R5 - it is still 1.0). I had a quick look at the Pax JDBC code. It doesn't appear to export this interface either. Perhaps it should (via the pax-jdbc module)? Anyway, I got around the problem (for now) by including the osgi.enterprise 4.2 jar. thanks again, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/org-osgi-service-jdbc-and-Karaf-tp4025573p4025577.html Sent from the Karaf - User mailing list archive at Nabble.com.
Karaf Cellar - Site Vs. Instance Configuration
Hello Jean-Baptiste (or anyone else on the Karaf team), From my understanding, cellar will automatically sync configuration between machines. Is my understanding correct? I also see the configuration files I added to the etc directory (these are all config files for declarative service apps) are owned by bundle Apache Karaf :: Cellar :: Config. Is that correct? If my understanding is correct, if I use Cellar what is the right way to do instance specific information (like IP addresses, instance ids etc.)? Should I be using files outside of configuration admin or is there another recommended way? While I am here, will Karaf 3 be coming out soon? No rush, just curious...:) thanks in advance, Gareth Collins -- View this message in context: http://karaf.922171.n3.nabble.com/Karaf-Cellar-Site-Vs-Instance-Configuration-tp3872157p3872157.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Wrong Execution Environment Names in Karaf 2.2.4?
Hello Jean-Baptiste, I second the request for JavaSE-1.6/JavaSE-1.7 execution environments. As well, I tried to add in the Equinox implementation of Declarative Services (unless I missed it, I didn't see a Karaf feature for Declarative Services). When I tried to install org.eclipse.equinox.util_1.0.200.v20100503.jar (required for installing equinox DS), it requested that execution environment OSGi/Minimum-1.1. I can go add this manually, but does that make sense to add this execution environment to the Karaf list as well? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Wrong-Execution-Environment-Names-in-Karaf-2-2-4-tp3447875p3451930.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: ActiveMq Listening on tcp6
Geoffry, The transport connectors configuration update is for your broker. Not for the broker clients (producer and consumer). regards, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/ActiveMq-Listening-on-tcp6-tp3334296p3336476.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Inconsistent deployment behavior between bundle installer and feature installer
Peter, I am curious - Which osgi container are you using with Karaf (Felix or Equinox)? If you are using Felix, could you try with Equinox and see if you see the same behaviour? thanks, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Inconsistent-deployment-behavior-between-bundle-installer-and-feature-installer-tp3322051p3336699.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: ActiveMq Listening on tcp6
Hello Geoffry, On the machine with your broker, do you see the following in netstat?: tcp0 0 :::server_ip:61616 :::* LISTEN tcp0 0 :::127.0.0.1:61616 :::* LISTEN If yes, from the consumer machine, can you now run: telnet server_ip 61616 and connect (to confirm the port is really accessible)? If yes, this is now an issue on your consumer. I am not understanding one thing though. How were you configuring your broker originally? I was expecting you to update an existing broker configuration file (whether it be outside OSGi, or defined inside OSGi as a spring-dm/blueprint file)...rather than creating a new one. regards, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/ActiveMq-Listening-on-tcp6-tp3334296p3337007.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: ActiveMq Listening on tcp6
Hello Geoffry, What you can put in the broker config comes from here: https://activemq.apache.org/schema/core/activemq-core-5.5.0.xsd To cut to the chase, I am attaching a broker blueprint file of mine which I know works. It is largely the default except for the additional transport connector and a network connector. Just remove the network connector (as you are not setting up a broker network yet) and change the IP address in the transport connector to reflect your producer IP and deploy to karaf. I am assuming the karaf instance which you will deploy to will have at least the activemq and activemq-blueprint features installed. Let me know if this works for you. regards, Gareth On Wed, Sep 14, 2011 at 4:38 PM, gcr [via Karaf] ml-node+s922171n3337079...@n3.nabble.com wrote: Gareth, This is what I get from netstat -anp --tcp: tcp6 0 0 127.0.0.1:61616 :::* LISTEN - It's that tcp6 business that concerns me. See below: On 14 September 2011 13:12, Gareth [hidden email] wrote: Hello Geoffry, On the machine with your broker, do you see the following in netstat?: tcp 0 0 :::server_ip:61616 :::* LISTEN tcp 0 0 :::127.0.0.1:61616 :::* LISTEN If yes, from the consumer machine, can you now run: telnet server_ip 61616 I'll have to install telnet first. and connect (to confirm the port is really accessible)? If yes, this is now an issue on your consumer. I am not understanding one thing though. How were you configuring your broker originally? I was expecting you to update an existing broker configuration file (whether it be outside OSGi, or defined inside OSGi as a spring-dm/blueprint file)...rather than creating a new one. I had a file called localhost-broker.xml that I created from the karaf client command line. (I don't recall the exact command I ran.) It contained a transport connector element similar to what you have prescribed. As I thrashed around, trying to get this thing to work, I removed it because none of the sketchy documentation I was referencing seemed to include it. regards, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/ActiveMq-Listening-on-tcp6-tp3334296p3337007.html Sent from the Karaf - User mailing list archive at Nabble.com. -- Geoffry Roberts If you reply to this email, your message will be added to the discussion below: http://karaf.922171.n3.nabble.com/ActiveMq-Listening-on-tcp6-tp3334296p3337079.html To unsubscribe from ActiveMq Listening on tcp6, click here. -- View this message in context: http://karaf.922171.n3.nabble.com/ActiveMq-Listening-on-tcp6-tp3334296p3337343.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Adding Additional Cellar Map Configuration?
Hello Ioannis, I believe this can be fixed by changing the HazelcastBundleListener to be a SynchronousBundleListener (I believe the error occurred because I accessed Hazelcast in my Activator.start()). Also I see the BundleEvent.STOPPING will remove the classloader. Will the bundle listener receive the STOPPING event before or after the Activator.stop is called (I would like to be able to execute a hazelcast action in Activator.stop)? If I run shutdown on karaf, it would be nice if it was possible to be able to execute a hazelcast command before cellar shuts down the hazelcast instance. When I try to do this now, it appears the cellar hazelcast bundle beats me to it (I get a hazelcast instance stopped error). Would you have any suggestion here? thanks in advance, Gareth On Wed, Sep 14, 2011 at 11:16 PM, Gareth Collins gareth.o.coll...@gmail.com wrote: Hello Ioannis, I tried my example again. I am still having a problem with the trunk cellar version (which should include the fix for 842). I am not sure the fix is completely there. The exception is different now (see below). thanks, Gareth java.lang.ClassNotFoundException: com.mytestcompany.PredTester at org.apache.karaf.cellar.core.utils.CombinedClassLoader.findClass(CombinedClassLoader.java:62) at java.lang.ClassLoader.loadClass(ClassLoader.java:306) at java.lang.ClassLoader.loadClass(ClassLoader.java:247) at java.lang.Class.forName0(Native Method) at java.lang.Class.forName(Class.java:247) at com.hazelcast.nio.AbstractSerializer.classForName(AbstractSerializer.java:72) at com.hazelcast.nio.AbstractSerializer.classForName(AbstractSerializer.java:57) at com.hazelcast.nio.Serializer$DataSerializer.classForName(Serializer.java:83) at com.hazelcast.nio.Serializer$DataSerializer.read(Serializer.java:93) at com.hazelcast.nio.Serializer$DataSerializer.read(Serializer.java:69) at com.hazelcast.nio.AbstractSerializer.toObject(AbstractSerializer.java:105) at com.hazelcast.nio.AbstractSerializer.toObject(AbstractSerializer.java:135) at com.hazelcast.nio.Serializer.readObject(Serializer.java:62) at com.hazelcast.impl.ThreadContext.toObject(ThreadContext.java:113) at com.hazelcast.nio.IOUtil.toObject(IOUtil.java:149) at com.hazelcast.impl.Record.getValue(Record.java:143) at com.hazelcast.query.Predicates$GetExpressionImpl.doGetValue(Predicates.java:842) at com.hazelcast.query.Predicates$GetExpressionImpl.getValue(Predicates.java:836) at com.hazelcast.query.Predicates$EqualPredicate.apply(Predicates.java:450) at com.hazelcast.query.PredicateBuilder.apply(PredicateBuilder.java:32) at com.hazelcast.impl.ConcurrentMapManager$QueryOperationHandler.createResultPairs(ConcurrentMapManager.java:2658) at com.hazelcast.impl.ConcurrentMapManager$QueryOperationHandler$QueryTask.run(ConcurrentMapManager.java:2627) at com.hazelcast.impl.executor.ParallelExecutorService$ParallelExecutorImpl$ExecutionSegment.run(ParallelExecutorService.java:179) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:680) On Wed, Sep 7, 2011 at 6:13 PM, iocanel [via Karaf] ml-node+3318095-519416612-416...@n3.nabble.com wrote: The branch 2.2.x is now back to stable. On Wed, Sep 7, 2011 at 10:28 AM, Ioannis Canellos [hidden email] wrote: You will need to wait a bit before you try 842. The branch needs some more fixes. -- Ioannis Canellos http://iocanel.blogspot.com Apache Karaf Committer PMC Apache ServiceMix Committer Apache Gora Committer -- Ioannis Canellos http://iocanel.blogspot.com Apache Karaf Committer PMC Apache ServiceMix Committer Apache Gora Committer -- Ioannis Canellos http://iocanel.blogspot.com Apache Karaf Committer PMC Apache ServiceMix Committer Apache Gora Committer Ioannis Canellos http://iocanel.blogspot.com If you reply to this email, your message will be added to the discussion below: http://karaf.922171.n3.nabble.com/Adding-Additional-Cellar-Map-Configuration-tp3305655p3318095.html To unsubscribe from Adding Additional Cellar Map Configuration?, click here. -- View this message in context: http://karaf.922171.n3.nabble.com/Adding-Additional-Cellar-Map-Configuration-tp3305655p3337894.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Cellar Commands Not Appearing In Karaf Console When Equinox Used
I should have checked JIRA. I see the commands in the trunk version. thanks again, Gareth On Sun, Sep 11, 2011 at 3:15 PM, Jean-Baptiste Onofré [via Karaf] ml-node+s922171n3327471...@n3.nabble.com wrote: You are right Achim ;) Regards JB On 09/11/2011 07:55 PM, Achim Nierbeck wrote: Hi Gareth, it's a known already fixed bug ;) https://issues.apache.org/jira/browse/KARAF-841 it will be available with cellar 2.2.3 Regards, Achim Am 11.09.2011 06:42, schrieb Gareth: Hello, I am having a small problem running cellar with equinox. If I start karaf 2.2.3 fresh (felix) and do the following: features:install webconsole features:addurl mvn:org.apache.karaf.cellar/apache-karaf-cellar/2.2.2/xml/features features:install cellar I see all the cellar commands appear in the karaf console (e.g. cluster:config-list, cluster:config-proplist etc). If I do the same with a fresh karaf 2.2.3 equinox (I set the karaf.framework to equinox in config.properties), the cellar commands do not appear (though cellar appears to still be running as my two Karaf nodes find each other). Anything obvious I missed that would cause the cellar commands not to appear in the console? Anything obvious I can check to help debug this? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Cellar-Commands-Not-Appearing-In-Karaf-Console-When-Equinox-Used-tp3326416p3326416.html Sent from the Karaf - User mailing list archive at Nabble.com. -- Jean-Baptiste Onofré [hidden email] http://blog.nanthrax.net Talend - http://www.talend.com If you reply to this email, your message will be added to the discussion below: http://karaf.922171.n3.nabble.com/Cellar-Commands-Not-Appearing-In-Karaf-Console-When-Equinox-Used-tp3326416p3327471.html To unsubscribe from Cellar Commands Not Appearing In Karaf Console When Equinox Used, click here. -- View this message in context: http://karaf.922171.n3.nabble.com/Cellar-Commands-Not-Appearing-In-Karaf-Console-When-Equinox-Used-tp3326416p3329766.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Adding Additional Cellar Map Configuration?
) ... 15 more thanks in advance, Gareth On Sun, Sep 11, 2011 at 3:49 AM, iocanel [via Karaf] ml-node+s922171n3326580...@n3.nabble.com wrote: It should be ok now. -- Ioannis Canellos FuseSource Blog: http://iocanel.blogspot.com Apache Karaf Committer PMC Apache ServiceMix Committer Apache Gora Committer Ioannis Canellos http://iocanel.blogspot.com If you reply to this email, your message will be added to the discussion below: http://karaf.922171.n3.nabble.com/Adding-Additional-Cellar-Map-Configuration-tp3305655p3326580.html To unsubscribe from Adding Additional Cellar Map Configuration?, click here. -- View this message in context: http://karaf.922171.n3.nabble.com/Adding-Additional-Cellar-Map-Configuration-tp3305655p3328350.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Adding Additional Cellar Map Configuration?
Hello Ioannis, I checked out the cellar-2.2.x branch: svn co http://svn.apache.org/repos/asf/karaf/cellar/branches/cellar-2.2.x and tried to do a build. It appears to be depending on 3.0.0-SNAPSHOT poms: [ERROR] The build could not read 1 project - [Help 1] org.apache.maven.project.ProjectBuildingException: Some problems were encountered while processing the POMs: [FATAL] Non-resolvable parent POM: Could not find artifact org.apache.karaf:cellar:pom:3.0.0-SNAPSHOT and 'parent.relativePath' points at wrong local POM @ line 24, column 13: . . [ERROR] [ERROR] The project org.apache.karaf.cellar:org.apache.karaf.cellar.dosgi:3.0.0-SNAPSHOT (/Users/gcollins/software/cellar-2.2.x/dosgi/pom.xml) has 1 error [ERROR] Non-resolvable parent POM: Could not find artifact org.apache.karaf:cellar:pom:3.0.0-SNAPSHOT and 'parent.relativePath' points at wrong local POM @ line 24, column 13 - [Help 2] org.apache.maven.model.resolution.UnresolvableModelException: Could not find artifact org.apache.karaf:cellar:pom:3.0.0-SNAPSHOT Did I miss something obvious? thanks in advance, Gareth On Wed, Sep 7, 2011 at 6:13 PM, iocanel [via Karaf] ml-node+3318095-519416612-416...@n3.nabble.com wrote: The branch 2.2.x is now back to stable. On Wed, Sep 7, 2011 at 10:28 AM, Ioannis Canellos [hidden email] wrote: You will need to wait a bit before you try 842. The branch needs some more fixes. -- Ioannis Canellos http://iocanel.blogspot.com Apache Karaf Committer PMC Apache ServiceMix Committer Apache Gora Committer -- Ioannis Canellos http://iocanel.blogspot.com Apache Karaf Committer PMC Apache ServiceMix Committer Apache Gora Committer -- Ioannis Canellos http://iocanel.blogspot.com Apache Karaf Committer PMC Apache ServiceMix Committer Apache Gora Committer Ioannis Canellos http://iocanel.blogspot.com If you reply to this email, your message will be added to the discussion below: http://karaf.922171.n3.nabble.com/Adding-Additional-Cellar-Map-Configuration-tp3305655p3318095.html To unsubscribe from Adding Additional Cellar Map Configuration?, click here. -- View this message in context: http://karaf.922171.n3.nabble.com/Adding-Additional-Cellar-Map-Configuration-tp3305655p3323570.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Configuring Cellar For TCP Instead Of Multicast
Hello Ioannis, I did try Jean-Baptiste's suggestion. I shut both instances down, added the tcpIpMembers to etc/org.apache.karaf.cellar.instance.cfg, started both instances up and it still didn't work. When I checked the etc/org.apache.karaf.cellar.instance.cfg configuration files, the tcpIpMembers field was empty: tcpIpMembers = Just to make sure I didn't miss something, I repeated the test with quotes around the tcpIpMembers: tcpIpMembers=192.168.204.123,192.168.204.124 with the same result. I haven't installed the cellar-cloud feature. I do notice another weird thing though...which I am not sure is related. I do see lots of duplicate fileinstall instances build up (see attached screenshot). I guess another minor point to all this. I will want to setup an apache activemq instance for each instance. The activemq instance will need different configuration for each instance. By default, when I create an activemq instance via activemq:create-broker a broker blueprint configuration file will be dropped in the deploy directory. Cellar aurtomatically tries to replicate anything in the deploy directory, doesn't it? What is the correct way of stopping it from doing this? Anyway, I guess I will try again with as few features installed as possible to see if I can get it to work. Any suggestions on what to try next would be much appreciated. thanks again, Gareth http://karaf.922171.n3.nabble.com/file/n3314332/Screen_shot_2011-09-06_at_1.53.40_PM.png -- View this message in context: http://karaf.922171.n3.nabble.com/Configuring-Cellar-For-TCP-Instead-Of-Multicast-tp3305497p3314332.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Configuring Cellar For TCP Instead Of Multicast
Hello Ioannis, I tried again with an even more basic configuration (base karaf + features:install cellar): State Version Name Repository Description [installed ] [1.9.3 ] hazelcastrepo-0 In memory data grid [installed ] [2.2.2 ] cellar repo-0 Karaf clustering [installed ] [1.0.0 ] jclouds repo-0 JClouds [installed ] [3.0] guicerepo-0 Google Guice [installed ] [3.0.6.RELEASE ] spring karaf-2.2.3 [installed ] [1.2.1 ] spring-dm karaf-2.2.3 [installed ] [2.2.3 ] config karaf-2.2.3 [installed ] [2.2.3 ] ssh karaf-2.2.3 [installed ] [2.2.3 ] management karaf-2.2.3 I still had the same problem. In fact, I found the problem even worse. As soon as you turn multicast off from the file and restart karaf, the cellar features bundle hangs on startup: [ 67] [Active ] [Created ] [ ] [ 60] Apache Karaf :: Cellar :: Core (2.2.2) [ 68] [Active ] [Created ] [ ] [ 60] Apache Karaf :: Cellar :: Config (2.2.2) [ 69] [Active ] [GracePeriod ] [ ] [ 60] Apache Karaf :: Cellar :: Features (2.2.2) [ 70] [Active ] [Created ] [ ] [ 60] Apache Karaf :: Cellar :: Bundle (2.2.2) [ 71] [Active ] [Created ] [ ] [ 60] Apache Karaf :: Cellar :: Utils (2.2.2) [ 72] [Active ] [Created ] [ ] [ 60] Apache Karaf :: Cellar :: Shell (2.2.2) [ 73] [Active ] [] [Started] [ 60] Apache Karaf :: Cellar :: Hazelcast (2.2.2) [ 74] [Active ] [Created ] [ ] [ 60] Apache Karaf :: Cellar :: Management (2.2.2) I have to shutdown karaf with kill -9 (as shutdown doesn't work). Even if I now turn multicast back on in the instance config file and restart, the Cellar :: Features bundle continues to hang. I don't see anything too exciting in the logs which could point me to the issue. I only see this on shutdown: 2011-09-06 15:29:23,259 | DEBUG | lixDispatchQueue | framework | ? ? | 0 - org.apache.felix.framework - 3.0.9 | BundleEvent STOPPED2011-09-06 15:29:23,259 | DEBUG | nt Dispatcher: 1 | BlueprintListener| raf.shell.osgi.BlueprintListener 85 | 30 - org.apache.karaf.shell.osgi - 2.2.3 | Blueprint app state changed to Destroying for bundle 74 2011-09-06 15:29:23,259 | WARN | hz.UDP.Sender| ManagementCenterService | dardLoggerFactory$StandardLogger 62 | - - | [cellar] sleep interrupted java.lang.InterruptedException: sleep interruptedat java.lang.Thread.sleep(Native Method)[:1.6.0_27] at com.hazelcast.impl.management.ManagementCenterService$UDPSender.run(ManagementCenterService.java:212)[42:hazelcast:1.9.3]2011-09-06 15:29:23,260 | DEBUG | FelixStartLevel | management | ? ? | 74 - org.apache.karaf.cellar.management - 2.2.2 | ServiceEvent UNREGISTERING 2011-09-06 15:29:23,261 | DEBUG | FelixStartLevel | ReferenceRecipe | eprint.container.ReferenceRecipe 152 | 9 - org.apache.aries.blueprint - 0.3.1 | Unbinding reference mbeanServer If you could let me know what else I can try, it would be much appreciated. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Configuring-Cellar-For-TCP-Instead-Of-Multicast-tp3305497p3314614.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Adding Additional Cellar Map Configuration?
Hello Ioannis, Making it possible to define the Hazelcast configuration would work. Great! I guess there may be be still one classloader issue with this distinct from issue 842. If you defined a map store in this configuration, I suspect hazelcast may look for this map store class before any bundles have imported the hazelcast service: hazelcast ... map name=myMap ... map-store enabled=true class-namecom.mytestcompany.examples.DummyStore/class-name write-delay-seconds0/write-delay-seconds /map-store /map /hazelcast Anyway, would you like me to give the 842 change a try? Is there an easy way to get the cellar 2.2.3 nightly (without having to build myself)? thanks again, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Adding-Additional-Cellar-Map-Configuration-tp3305655p3314719.html Sent from the Karaf - User mailing list archive at Nabble.com.
Configuring Cellar For TCP Instead Of Multicast
| MessageDatabase
| emq.store.kahadb.MessageDatabase 1280 | 81 -
org.apache.activemq.activemq-core - 5.5.0 | Checkpoint done.
2011-09-02 18:36:20,371 | INFO | .1.ServiceThread | ClusterManager
| dardLoggerFactory$StandardLogger 62 | - - | [cellar]
Members [2] {
Member [192.168.204.123:5701] this
Member [192.168.204.124:5701]
}
2011-09-02 18:36:22,264 | DEBUG | heckpoint Worker | MessageDatabase
| emq.store.kahadb.MessageDatabase 1161 | 81 -
org.apache.activemq.activemq-core - 5.5.0 | Checkpoint started.
2011-09-02 18:36:22,268 | DEBUG | heckpoint Worker | MessageDatabase
| emq.store.kahadb.MessageDatabase 1280 | 81 -
org.apache.activemq.activemq-core - 5.5.0 | Checkpoint done.
2011-09-02 18:36:22,377 | DEBUG | hz.1.InThread| Connection
| dardLoggerFactory$StandardLogger 62 | - - | [cellar] Connection lost
/192.168.204.124:34112
2011-09-02 18:36:22,377 | WARN | hz.1.InThread| ReadHandler
| dardLoggerFactory$StandardLogger 62 | - - | [cellar] hz.1.InThread
Closing socket to endpoint Address[192.168.204.124:5701],
Cause:java.io.EOFException
2011-09-02 18:36:22,377 | INFO | .1.ServiceThread | ClusterManager
| dardLoggerFactory$StandardLogger 62 | - - | [cellar] Removing Address
Address[192.168.204.124:5701]
2011-09-02 18:36:22,382 | INFO | .1.ServiceThread | ClusterManager
| dardLoggerFactory$StandardLogger 62 | - - | [cellar]
Members [1] {
Member [192.168.204.123:5701] this
}
2011-09-02 18:36:22,390 | INFO | hz.1.InThread| InSelector
| dardLoggerFactory$StandardLogger 62 | - - | [cellar] 5701 is
accepting socket connection from /192.168.204.124:38994
2011-09-02 18:36:22,391 | INFO | hz.1.InThread| InSelector
| dardLoggerFactory$StandardLogger 62 | - - | [cellar] 5701 is accepted
socket connection from /192.168.204.124:38994
I am sure I have missed something obvious here (though I haven't found it
yet in the docs). Anything obvious I have messed up?
thanks in advance,
Gareth
--
View this message in context:
http://karaf.922171.n3.nabble.com/Configuring-Cellar-For-TCP-Instead-Of-Multicast-tp3305497p3305497.html
Sent from the Karaf - User mailing list archive at Nabble.com.
Adding Additional Cellar Map Configuration?
Hello all, Is there any way to add additional Cellar/Hazelcast configuration? For example, if I reuse the existing Hazelcast service I may want to add persistence to my Maps (this will be another classloader problem as I would need to define my MapStore implementation in the configuration)...or I may want to use a Hazelcast map as a cache with an eviction policy...or it may be useful to define a near cache. Is this possible now? If not, could this feature be added to the roadmap for Cellar? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Adding-Additional-Cellar-Map-Configuration-tp3305655p3305655.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Cellar And Hazelcast Questions
Hello Ioannis, Thank you very much for all your responses. It has been very helpful. Please bear with me. I must be missing something fundamental here. I realize there are two variables in play here rather than one (i.e. who initializes the Hazelcast instance, whether the TCCL is used). I don't believe Hazelcast can inherit my TCCL if my bundle only uses Hazelcast as a service, can it? When you ran your test, was the Hazelcast instance initialized as part of the same bundle? I did run through a few scenarios to confirm my understanding: (1) Use existing Karaf Hazelcast instance, use TCCL - FAILED (2) Create my own Hazelcast instance, don't use TCCL - FAILED (3) Create my own Hazelcast instance, use TCCL - SUCCESS! So it appears that unless I create my own Hazelcast instance, Hazelcast cannot see my classes (unless I use fragments, of course). Have I missed something obvious here? Would it be an idea to get cellar to use my hazelcast instance? Thus in my bundle which starts hazelcast, I could make sure I import both cellar and my own packages. Thus hazelcast can load both cellar and my classes via the TCCL. Would that make sense? Thank you very much again for all your help. regards, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Cellar-And-Hazelcast-Questions-tp3184320p3294046.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Cellar And Hazelcast Questions
Hello Ioannis, I was wondering. Would using fragments be a reasonable way to solve Hazelcast classloading issues (i.e. any class which may need to be serialized/deserialized by Hazelcast, I should include in a fragment which is attached to the Hazelcast bundle)? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Cellar-And-Hazelcast-Questions-tp3184320p3276278.html Sent from the Karaf - User mailing list archive at Nabble.com.
Assigning Wars To Ports And Subdirectories?
Hello, I am not sure whether this is a PAX Web question, or a Karaf question but: (1) If I install a war/wab file, is there any way of making it available on certain ports? For example, I would probably want the Karaf console and Active MQ console on a separate port to my application wabs/wars (as my wabs/wars may be visible through a firewall whilst Karaf console should not be). (2) Any way of installing wars (which are not OSGi wabs) to subdirectories? In tomcat I could do this by adding a # as part of the war name (e.g. hosting#fred.war could be accessed by http://hostname/hosting/fred.war). There isn't a similar feature in Karaf/PAX Web, is there? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Assigning-Wars-To-Ports-And-Subdirectories-tp3210710p3210710.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Assigning Wars To Ports And Subdirectories?
Hello JB, I knew about the Web-ContextPath in the MANIFEST.MF. I was just hoping there was a way to be able to do this without making all wars into wabs :). For these wars, could I perhaps create my own bundle listener, make it run before Pax Web (is that possible?) and go add the Web-ContextPath into the MANIFEST.MF before PAX Web reads it (is that possible?)? I also gave your two suggestions a try: (1) With the first suggestion, I got more ports configured for Jetty, but all my wars were still available on all ports. (2) I added the file install configuration (I see now two file install configurations in the Karaf console). I can now add files from the new location...though web applications are still installed in the same dir. There is no way to run multiple Pax Web instances in the same Karaf instance, is there (with the ability to explicitly assign each war to a specific instance?)? Thank you very much for the help. regards, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Assigning-Wars-To-Ports-And-Subdirectories-tp3210710p3211059.html Sent from the Karaf - User mailing list archive at Nabble.com.
Using Spring DM For New Software?
Hello, In a earlier question today I mentioned that when I tried to use Spring security in OSGi, I had to include all the Spring + spring security jars in my wab, which was not ideal. I found out what the problem was. I had to include Spring DM web to be able to remove the jars. The following was added to my web.xml: context-param param-namecontextClass/param-name param-value org.springframework.osgi.web.context.support.OsgiBundleXmlWebApplicationContext /param-value /context-param Since this is new code, I was a little uncomfortable adding a dependency on something which was essentially EOL so I had a look to see if I could do the same with Blueprint (as we are already planning to use blueprint for other non-web bundles). Unfortunately the Spring DM web functionality doesn't appear to be covered in the Blueprint spec (and I don't see anything currently in Aries). So my question is - is there another method for including Spring in a wab? If not and I continue with Karaf, what is the recommended approach going forward? Continue using Spring DM web for the foreseeable future? Avoid creating dependencies on Spring DM web and include the Spring jars in every wab? Any other options in the pipeline (will Aries eventually have a similar feature, or would a replacement come as part of PAX Web?)? Any suggestions/info/guidance would be much appreciated. thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Using-Spring-DM-For-New-Software-tp3205466p3205466.html Sent from the Karaf - User mailing list archive at Nabble.com.
Re: Cellar And Hazelcast Questions
Hello Ioannis,
I found where the issue occurs. It occurs when I try and execute a predicate
on a class I defined. Here is the sample code I used to reproduce the
problem:
My custom class:
public class PredTester implements DataSerializable {
private static final long serialVersionUID = -1359492685240522455L;
private String myvalue;
public PredTester(String value) {
this.myvalue = value;
}
@Override
public void writeData(DataOutput out) throws IOException {
out.writeUTF(myvalue);
}
public String getMyvalue() {
return this.myvalue;
}
@Override
public void readData(DataInput in) throws IOException {
this.myvalue = in.readUTF();
}
}
My activator:
public class Activator implements BundleActivator {
private volatile HazelcastInstance instance = null;
private volatile Thread myThread = null;
private volatile IMaplt;String,PredTestergt; myMap = null;
@Override
public void start(BundleContext context) throws Exception {
ClassLoader classLoader =
Thread.currentThread().getContextClassLoader();
Thread.currentThread().setContextClassLoader(this.getClass().getClassLoader());
ServiceReference reference =
context.getServiceReference(com.hazelcast.core.HazelcastInstance);
instance = (HazelcastInstance) context.getService(reference);
context.ungetService(reference);
try {
myMap = instance.getMap(fred-osgi);
myMap.addIndex(myvalue, true);
myMap.put(test1, new PredTester(belt));
EntryObject e = new PredicateBuilder().getEntryObject();
@SuppressWarnings(unchecked)
Predicatelt;String,PredTestergt; predicate =
e.get(myvalue).equal(belt);
Setlt;Map.Entrylt;String,PredTestergt; entrySet =
myMap.entrySet(predicate);
logger.info(Entries returned = {},entrySet.size()); I
believe it
fails here
} catch (Exception ex) {
ex.printStackTrace();
}
Thread.currentThread().setContextClassLoader(classLoader);
}
The following exception is thrown, whether I use an index or not (I thought
using an index may stop Hazelcast needing to deserialize to check for a
match):
21:01:16,037 | ERROR | .cached.thread-1 | AbstractSerializer |
dardLoggerFactory$StandardLogger 58 | - - | Problem reading
DataSerializable class : PredTester, exception:
java.lang.ClassNotFoundException: PredTester not found from bundle
[org.apache.karaf.cellar.hazelcast]
java.io.IOException: Problem reading DataSerializable class :
com.antennasoftware.hazelcasttest.PredTester, exception:
java.lang.ClassNotFoundException:
com.antennasoftware.hazelcasttest.PredTester not found from bundle
[org.apache.karaf.cellar.hazelcast]
at
com.hazelcast.nio.Serializer$DataSerializer.read(Serializer.java:98)[135:hazelcast:1.9.3]
at
com.hazelcast.nio.Serializer$DataSerializer.read(Serializer.java:69)[135:hazelcast:1.9.3]
at
com.hazelcast.nio.AbstractSerializer.toObject(AbstractSerializer.java:105)[135:hazelcast:1.9.3]
at
com.hazelcast.nio.AbstractSerializer.toObject(AbstractSerializer.java:135)[135:hazelcast:1.9.3]
at
com.hazelcast.nio.Serializer.readObject(Serializer.java:62)[135:hazelcast:1.9.3]
at
com.hazelcast.impl.ThreadContext.toObject(ThreadContext.java:113)[135:hazelcast:1.9.3]
at
com.hazelcast.nio.IOUtil.toObject(IOUtil.java:149)[135:hazelcast:1.9.3]
at
com.hazelcast.impl.Record.getValue(Record.java:143)[135:hazelcast:1.9.3]
at
com.hazelcast.query.Predicates$GetExpressionImpl.doGetValue(Predicates.java:842)[135:hazelcast:1.9.3]
at
com.hazelcast.query.Predicates$GetExpressionImpl.getValue(Predicates.java:836)[135:hazelcast:1.9.3]
at
com.hazelcast.query.Predicates$EqualPredicate.apply(Predicates.java:450)[135:hazelcast:1.9.3]
at
com.hazelcast.query.PredicateBuilder.apply(PredicateBuilder.java:32)[135:hazelcast:1.9.3]
at
com.hazelcast.impl.ConcurrentMapManager$QueryOperationHandler.createResultPairs(ConcurrentMapManager.java:2658)[135:hazelcast:1.9.3]
at
com.hazelcast.impl.ConcurrentMapManager$QueryOperationHandler$QueryTask.run(ConcurrentMapManager.java:2627)[135:hazelcast:1.9.3]
at
com.hazelcast.impl.executor.ParallelExecutorService$ParallelExecutorImpl$ExecutionSegment.run(ParallelExecutorService.java:179)[135:hazelcast:1.9.3]
at
java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)[:1.6.0_26]
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)[:1.6.0_26]
at java.lang.Thread.run(Thread.java:680)[:1.6.0_26]
I guess this makes it difficult to use Hazelcast predicates in OSGi (for
now).
thanks again,
Gareth
Re: Cellar And Hazelcast Questions
Hello JB, Thank you very much for responding so quickly! I am just using the default cluster group (only one machine currently as I am testing on my Mac): karaf@root cluster:group-list Node Group * 192.168.190.1:5701 default I am getting the HazelcastInstance via the service cellar creates (directly now as I am still exploring OSGi, probably using blueprint in the future): ServiceReference reference = bundleContext.getServiceReference(com.hazelcast.core.HazelcastInstance); HazelcastInstance instance = (HazelcastInstance) bundleContext.getService(reference); bundleContext.ungetService(reference); Then I am looking up/creating two maps (MyKeyType and MyValueType are defined in the same bundle - the exception thrown below is for MyKeyType): IMaplt;String,Longgt; myMap1 = hazelcastInstance.getMap(myMap1); IMaplt;MyKeyType,MyValueTypegt; myMap2 = hazelcastInstance.getMap(myMap2); Is that enough information? Am I doing anything I shouldn't be doing? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Cellar-And-Hazelcast-Questions-tp3184320p3185631.html Sent from the Karaf - User mailing list archive at Nabble.com.
Cellar And Hazelcast Questions
Hello, I would like to use Karaf as well as Cellar. I also would like to use Hazelcast for my program (since it is now getting a lot of Apache integration vs. other open-source clustering software). So I have some questions: (1) Is there an issue if I piggyback on the Cellar Hazelcast cluster, or should I make a completely separate one? Just as a note, I am seeing some exceptions when I uninstall my bundles after using Hazelcast: java.lang.ClassNotFoundException:my class not found from bundle [org.apache.karaf.cellar.hazelcast] at org.springframework.osgi.util.BundleDelegatingClassLoader.findClass(BundleDelegatingClassLoader.java:103) at org.springframework.osgi.util.BundleDelegatingClassLoader.loadClass(BundleDelegatingClassLoader.java:156) at java.lang.ClassLoader.loadClass(ClassLoader.java:247) at java.lang.Class.forName0(Native Method) at java.lang.Class.forName(Class.java:247) . . Caused by: java.lang.ClassNotFoundException: My class not found by org.apache.karaf.cellar.hazelcast [141] at org.apache.felix.framework.ModuleImpl.findClassOrResourceByDelegation(ModuleImpl.java:787) at org.apache.felix.framework.ModuleImpl.access$400(ModuleImpl.java:71) at org.apache.felix.framework.ModuleImpl$ModuleClassLoader.loadClass(ModuleImpl.java:1768) at java.lang.ClassLoader.loadClass(ClassLoader.java:247) at org.apache.felix.framework.ModuleImpl.getClassByDelegation(ModuleImpl.java:645) (2) I see there is an OSGi classloader problem with the Hazelcast Cluster Monitor. If you use any custom classes in an IMap it cannot find them. e.g.: Caused by: java.lang.ClassNotFoundException: org.apache.karaf.cellar.core.Group not found by com.hazelcast.monitor [143] at org.apache.felix.framework.ModuleImpl.findClassOrResourceByDelegation(ModuleImpl.java:787) at org.apache.felix.framework.ModuleImpl.access$400(ModuleImpl.java:71) at org.apache.felix.framework.ModuleImpl$ModuleClassLoader.loadClass(ModuleImpl.java:1768) at java.lang.ClassLoader.loadClass(ClassLoader.java:247)[:1.6.0_26] Is there any way around this? Is using the Hazelcast cluster monitor to be avoided for OSGi/Keraf Cellar? thanks in advance, Gareth -- View this message in context: http://karaf.922171.n3.nabble.com/Cellar-And-Hazelcast-Questions-tp3184320p3184320.html Sent from the Karaf - User mailing list archive at Nabble.com.
