Timeout when starting JobManager within Eclipse for debugging purposes
Locked
Timeout when starting JobManager within Eclipse for debugging purposes
 
|
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1 Hi everybody, I'd like to start the JobManager in debug mode from within Eclipse. I used "--configDir PATH_TO_CONFDIR --executionMode local" as parameters. In earlier versions that was enough. However, now I get a timeout exeption that stems from the akka actor system. I am using the latest snapshot build 0.9 and the stack trace is at the bottom of the mail. Am I missing something? Do I need to provide another parameter? Thanks a lot and regards Mathias Stack: Exception in thread "main" java.util.concurrent.TimeoutException: Futures timed out after [60000 milliseconds] at scala.concurrent.impl.Promise$DefaultPromise.ready(Promise.scala:219) at scala.concurrent.impl.Promise$DefaultPromise.result(Promise.scala:223) at scala.concurrent.Await$$anonfun$result$1.apply(package.scala:107) at scala.concurrent.BlockContext$DefaultBlockContext$.blockOn(BlockContext.scala:53) at scala.concurrent.Await$.result(package.scala:107) at akka.remote.Remoting.start(Remoting.scala:180) at akka.remote.RemoteActorRefProvider.init(RemoteActorRefProvider.scala:184) at akka.actor.ActorSystemImpl.liftedTree2$1(ActorSystem.scala:619) at akka.actor.ActorSystemImpl._start$lzycompute(ActorSystem.scala:616) at akka.actor.ActorSystemImpl._start(ActorSystem.scala:616) at akka.actor.ActorSystemImpl.start(ActorSystem.scala:633) at akka.actor.ActorSystem$.apply(ActorSystem.scala:142) at akka.actor.ActorSystem$.apply(ActorSystem.scala:119) at akka.actor.ActorSystem$.create(ActorSystem.scala:67) at org.apache.flink.runtime.akka.AkkaUtils$.createActorSystem(AkkaUtils.scala:52) at org.apache.flink.runtime.akka.AkkaUtils$.createActorSystem(AkkaUtils.scala:40) at org.apache.flink.runtime.jobmanager.JobManager$.main(JobManager.scala:464) at org.apache.flink.runtime.jobmanager.JobManager.main(JobManager.scala) -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJUqSscAAoJEHj0VWOtoCa75ZQH/0C58ucPLu8z9f0pgiGGsRfJ 4+xsNXRjyepoPk9hFAcmVV4C0uDBW+F24rVJF7ZVA7lLWiDgy6sydLZXoKBgOXZm Kgq2NrFvPzlFdIPs1tu885SDr8rKAw6tYhm21NDALb7kPNZGjtJv1QxgMN72n7vh hkdzu4VqVut0qh8hPY7mdChIs4zxERoytg92vLv/oJ84nmsHtnOmB+SRwLNsO1d5 zOqBd+I31bJGsB1TY+wtYc+AkqfMK1vMmYLqCgCW/vu6GWpvNOpljYic0vNDhkQU 0i40uDcn4VVDONsyQhYXx3XbWyTjF2jr83tNyU4AKZklNIV1xJMQBZMvF9sZuXU= =C3/B -----END PGP SIGNATURE----- |
Re: Timeout when starting JobManager within Eclipse for debugging purposes
|
|
Hi Mathias, the timeout exception means that the actor system could not be started. Usually this is the case if the hostname is not specified or if the hostname cannot be looked up. The best way to see what is happening is to turn on logging by providing a log4j.properties file with DEBUG logging. Additionally you can turn on akka logging by inserting akka.loglevel: DEBUG into the config file. A stripped down log4j config file looks the following: log4j.rootLogger=DEBUG, A1 log4j.appender.A1=org.apache.log4j.ConsoleAppender log4j.appender.A1.layout=org.apache.log4j.PatternLayout log4j.appender.A1.layout.ConversionPattern=%-4r [%t] %-5p %c %x - %m%n I just tried to reproduce your problem on my machine, but I did not succeed. I started the JobManager with the config values jobmanager.rpc.address: localhost akka.loglevel: DEBUG For debugging in general, it is important that now all blocking RPC calls have now a timeout with the latest Akka changes. Thus, by blocking or slowing down a receiving thread too long might trigger a timeout exception on the receiving side. One can solve this problem by setting akka.ask.timeout in the config to a high value. I'll also increase the default timeout for local execution in one of the next commits. If you still have the debugging problem, could you then send me the log output. Greets, Till On Sun, Jan 4, 2015 at 12:59 PM, Mathias Peters <[hidden email]> wrote: -----BEGIN PGP SIGNED MESSAGE----- |
Re: Timeout when starting JobManager within Eclipse for debugging purposes
|
|
I forgot to mention that the parameters you used are perfectly correct. On Mon, Jan 5, 2015 at 4:57 PM, Till Rohrmann <[hidden email]> wrote:
|
Re: Timeout when starting JobManager within Eclipse for debugging purposes
|
|
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1 Hi Till, turned out I chose the wrong confDir-Path. The directory didn't contain the flink-conf.yaml. ;) AkkaUtils tried to create the ActorSystem anyway so I didn't realize at first. I suggest a warning in case the conf files cannot be found. Thanks for the reply! Regards Mathias > On 05.01.2015 16:59, Till Rohrmann wrote: >> I forgot to mention that the parameters you used are perfectly >> correct. > >> On Mon, Jan 5, 2015 at 4:57 PM, Till Rohrmann >> <[hidden email]> wrote: > >>> Hi Mathias, >>> >>> the timeout exception means that the actor system could not be >>> started. Usually this is the case if the hostname is not >>> specified or if the hostname cannot be looked up. The best way >>> to see what is happening is to turn on logging by providing a >>> log4j.properties file with DEBUG logging. Additionally you can >>> turn on akka logging by inserting akka.loglevel: DEBUG into >>> the config file. A stripped down log4j config file looks the >>> following: >>> >>> log4j.rootLogger=DEBUG, A1 >>> log4j.appender.A1=org.apache.log4j.ConsoleAppender >>> log4j.appender.A1.layout=org.apache.log4j.PatternLayout >>> log4j.appender.A1.layout.ConversionPattern=%-4r [%t] %-5p %c %x >>> - %m%n >>> >>> I just tried to reproduce your problem on my machine, but I >>> did not succeed. I started the JobManager with the config >>> values >>> >>> jobmanager.rpc.address: localhost akka.loglevel: DEBUG >>> >>> For debugging in general, it is important that now all >>> blocking RPC calls have now a timeout with the latest Akka >>> changes. Thus, by blocking or slowing down a receiving thread >>> too long might trigger a timeout exception on the receiving >>> side. One can solve this problem by setting akka.ask.timeout in >>> the config to a high value. I'll also increase the default >>> timeout for local execution in one of the next commits. >>> >>> If you still have the debugging problem, could you then send >>> me the log output. >>> >>> Greets, >>> >>> Till >>> >>> On Sun, Jan 4, 2015 at 12:59 PM, Mathias Peters >>> <[hidden email]> wrote: >>> >> Hi everybody, > >> I'd like to start the JobManager in debug mode from within >> Eclipse. I used "--configDir PATH_TO_CONFDIR --executionMode >> local" as parameters. In earlier versions that was enough. >> However, now I get a timeout exeption that stems from the akka >> actor system. > >> I am using the latest snapshot build 0.9 and the stack trace is >> at the bottom of the mail. > >> Am I missing something? Do I need to provide another parameter? > >> Thanks a lot and regards Mathias > > >> Stack: Exception in thread "main" >> java.util.concurrent.TimeoutException: Futures timed out after >> [60000 milliseconds] at >> scala.concurrent.impl.Promise$DefaultPromise.ready(Promise.scala:219) > >> > > at >> scala.concurrent.impl.Promise$DefaultPromise.result(Promise.scala:223) > >> > > at >> scala.concurrent.Await$$anonfun$result$1.apply(package.scala:107) >> at > >> scala.concurrent.BlockContext$DefaultBlockContext$.blockOn(BlockContext.scala:53) > >> > > at scala.concurrent.Await$.result(package.scala:107) >> at akka.remote.Remoting.start(Remoting.scala:180) at >> akka.remote.RemoteActorRefProvider.init(RemoteActorRefProvider.scala:184) > >> > > at akka.actor.ActorSystemImpl.liftedTree2$1(ActorSystem.scala:619) >> at >> akka.actor.ActorSystemImpl._start$lzycompute(ActorSystem.scala:616) > >> > > at akka.actor.ActorSystemImpl._start(ActorSystem.scala:616) >> at akka.actor.ActorSystemImpl.start(ActorSystem.scala:633) at >> akka.actor.ActorSystem$.apply(ActorSystem.scala:142) at >> akka.actor.ActorSystem$.apply(ActorSystem.scala:119) at >> akka.actor.ActorSystem$.create(ActorSystem.scala:67) at > >> org.apache.flink.runtime.akka.AkkaUtils$.createActorSystem(AkkaUtils.scala:52) > >> > > at > >> org.apache.flink.runtime.akka.AkkaUtils$.createActorSystem(AkkaUtils.scala:40) > >> > > at >> org.apache.flink.runtime.jobmanager.JobManager$.main(JobManager.scala:464) > >> > > at >> org.apache.flink.runtime.jobmanager.JobManager.main(JobManager.scala) > >> > >>>> >>> >>> > > > -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQEcBAEBAgAGBQJUq5DcAAoJEHj0VWOtoCa7vPgIAN0Kif/27LNfq4Un0QD4klY2 nvsC3ulM2/AS0IgRYLzN4mJBCMV2+GzsA2rqbnTNzh6CfscYK3ij+TLUGt5+Ttju wwtfOjiSgri5ySVuiITHHb4OpTG+ZUi0d6pHZ9KVBf82Z98tHDzxncaqn1vN1r1a wFLLUprDfjKlOBMaQ8lXrNj+ysglpuLN33IFv2evW8ZliaPZbnj6JroB+SETotLO OWw7YBHGRqPDcQmXh/fOkFT6PryN63OgQ89yB9oIoLp45RNmDOm0RILzKWXaM9Ah eOAVJ9ln7MHeyxCM6e2dQz3FOGFC81F6BTW0vFXLBedGAcDZKiG18jvNWjABQJ0= =fcsi -----END PGP SIGNATURE----- |
Re: Timeout when starting JobManager within Eclipse for debugging purposes
|
|
Mathias, That makes total sense. Do you want to open an issue for that? In addition: what prevents us from allowing to start a local mode job manager without a config? Stephan Am 06.01.2015 08:38 schrieb "Mathias Peters" <[hidden email]>:
-----BEGIN PGP SIGNED MESSAGE----- |
Re: Timeout when starting JobManager within Eclipse for debugging purposes
|
|
In reply to this post by Mathias Peters
Ah good to know. Yeah you're right, it does not make much sense to create a remote actor system where the hostname is null or empty. I'll add a fix for that. Greetings, Till On Tue, Jan 6, 2015 at 8:38 AM, Mathias Peters <[hidden email]> wrote: -----BEGIN PGP SIGNED MESSAGE----- |
Re: Timeout when starting JobManager within Eclipse for debugging purposes
|
|
In reply to this post by Stephan Ewen
Sorry I saw the follow up mail too late. There is no reason why we cannot start the job manager in local mode without a configuration. In fact, with the latest refactoring it does no longer require a hostname for the local mode. Greets, Till On Tue, Jan 6, 2015 at 9:00 AM, Stephan Ewen <[hidden email]> wrote:
|
«
Return to (DEPRECATED) Apache Flink User Mailing List archive.
|
1 view|%1 views
| Free forum by Nabble | Edit this page |