Terracotta Discussion Forums (LEGACY READ-ONLY ARCHIVE)

Hello,
It means that the Management Server, the code that provides a REST API for management and monitoring of your cache managers, found more than 1 file named META-INF/services/net.sf.ehcache.management.ManagementServer in your application classpath - and it's not supposed to happen (it uses underneath a ServiceLoader that determines whether to use the OpenSOurce or Enterprise code)

in a nutshell, it means you have classpath issues : I'd say you have several ehcache jars in your classpath.

If you could list all the jars found in your classpath, and if you provided the version of ehcache embedded, we could help you some more.

And if you are not interested in the Rest API for management and monitoring, you can simply disable it and happily forget this message : in your ehcache.xml :
<managementRESTService enabled="false"/> below your <ehcache ..> tag

Hello,
I could make recovery work with this code :

Code:

JobDetail job1 = newJob(MyJob.class).requestRecovery().withIdentity("job1", "group1").build();

and a similar quartz configuration.
Quartz will try to recover jobs that did not complete, take this as an example :
Code:

public void execute(JobExecutionContext context)
       throws JobExecutionException {
 
     if (context.isRecovering()) {
       System.err.println("Job tried to recover !");
       _log.info("My thread is :" + Thread.currentThread().getName());
     }
     _log.info("Get ready to die");
     if (shouldDie()) {
       try {
         Runtime.getRuntime().halt(255);
       } catch (Exception e1) {
         // TODO Auto-generated catch block
         e1.printStackTrace();
       }
     }
   }

hello,
yes if your job is interrupted and you request receovery, the next time you start your app you should see a recovering trigger firing your job.
Are you using quartz connected to a database ?

have you tried checking the code out from http://svn.terracotta.org/svn/quartz/branches/2.2.1-rc/ ?

hello,
in your use case, your webclients are going to use (embed) ehcache as the terracotta server array clients.
Your webclients code will interact with ehcache to do the puts/gets/deletes on the saved data on the TSA.

hit ratio , hit rate are "live" /dynamic values
don't expect to see anything if nothing is going on

hello,
what I mean is that I'd like to make sure that there are hits and misses on your caches before investigating why the ehcache rest agent reported almost no activity.
I think in your case everything's working normally : your app (through hibernate) does not seem to be hitting the caches.
For the reproducible example, I'd like the simplest reproducible test case; so that I'm sure to reproduce it.
thanks

Hello,
those responses you got indicate that there are few hits on the caches.
Did you make sure there is activity wrt the cache ? (frequent get/put)
I'm not talking about hibernate here, the agent will only give you cache activity.
If you're sure the agent did not report proper values, please share a sample project so that we can re run it ourselves and reproduce the issue.
thanks

Hello,
I could not reproduce your error, even specifying a license path with -Dcom.tc.productkey.path=/path/to/terracotta-license.key
That said, I acknowledge that the Group.class does not exist, and that the ehcache and slf4j jars also need to be added to your classpath (I have asked the webmasters to update the website accordingly)

Here is a screenshot I took of my workspace : no need to specify any vm argument for the license if you follow my setup.

https://dl.dropboxusercontent.com/u/2138522/bigmemory-max-tryout.png

I hope that helps, and thanks for letting us know !
Anthony

Hello,
A few ideas :
* can you try to change the polling setting to 5 sec ? and check if the values are not null anymore ?
* can you try, in your browser, to refresh a few time this resource (the Ehcache Rest agent response) :
http://localhost:9888/tc-management-api/agents/cacheManagers/caches/statistics/samples

--> if hitratio or hit rate returning 0 and should not, please tell us and send us your project (or a similar one reproducing the issue if you can't)
Here is what I get locally : http://pastie.org/8372134

* can you try, in your browser to refresh a few time this resource (the TMC Rest Api aggregated response) :
http://localhost:9889/tmc/api/agents/cacheManagers/caches/statistics/samples

--> if hitratio or hit rate returning 0 and should not, please tell us and send us your project (or a similar one reproducing the issue if you can't)
Here is what I get locally : http://www.pastebin.ca/2461167

With those results, I'll be able to determine where your issue is : on the Ehcache Agent side or on the TMC side

Thank you

hello,
could you please provide us with more details about your project ?
How did you set the classpath (maven, ant or just adding to the IDE classpath ?)
Even better : can you zip your project and attach it to the forum post ? (you would not need to include all the big memory jars, just tell us which jars , with precise version, you relied on)
thank you in advance

Hello Amber,
Thanks for sending your configuration.
Your configuration looks valid regarding clustering.
Could you please give me the responses of such requests (in curl or a web browser) :

http://localhost:9889/tmc/api/agents

http://localhost:9889/tmc/api/agents/info

http://XXX.droot.net:9530/tc-management-api/agents

http://XXX.droot.net:9530/tc-management-api/agents/info

The 2 first ones are really important : is your cache manager connected to the TSA or not ?
Thanks,
Anthony

Hello Amber,
Thanks for attaching the log.
The logs do not report any error, since you left configured the default connections (LocalConnectionGroup listening on localhost:9888 and LocalCluster listenign on localhost:9530) and added a new Cluster connection.
What I think is that your CacheManager was not configured to connect to your cluster (ie : there are no clients to the TSA)
Could you please make sure your ehcache configuration is configured to connect to the cluster (please attanch your ehcache configuration ehcache.xml)
thanks

Hello Amber,

amberg wrote:

1. Once a new connection is created for monitoring then I am starting a client to create a cache manager and store data. Unless I stop the client I cannot see client details. I am interested in observing hit ratios while client is running.

Yes, of course you should be able to see hit ratios and other statistics !
I assume you defined a clustered cache, and that the Terracotta Server Array you connected your clustered cache to is monitored by the Terracotta Management Console.
I also assume you waited at least few seconds for statistics to appear.
Now, I suggest you attach the logs of the tmc (found under bigmemory-max-4.0.2/tools/management-console/logs/xxx.stderrout.log) to this forum post, so that we can investigate what went wrong

amberg wrote:

2. Under Appication data there are no cache managers visible and a error message is displayed as "There are no Cache Managers" I understood cache manager as created by client application should be visible on this tab.

right , this is the same issue as 1.
thanks,
Anthony

Hello Zack,
Statistics still work in big memory 4.0.2, both using the api or the rest agent and the tms.
Could you please share a minimal test that reproduces the issue ? (with configuration and code that "does something" to the cache) ?
Thanks in advance,
Anthony