i implemented what your wrote in the link but got in the storm ui zeros in emitted and transfered also have in the supervisor log file couldn't extract resources from jar file then got still hasn't start
i got at hte end of the drpc log file Starting Distributed RPC servers... On Thu, Dec 24, 2015 at 9:27 PM, researcher cs <[email protected]> wrote: > many thanks for your answers , sorry for my lot question but i need to > understand many thanks for you > > Now i tested the topology bu this command > > 1-mvn package > mvn compile exec:java -Dexec.classpathScope=compile > -Dexec.mainClass=trident.Topology > > that mean i tested it in local rigth ? > > > 2- now i want to implement 2 point which coder wrote > "production mode " > you said that i should submit the topology first in single machine with > the above configuration right ? > > 3- should i write myipaddress started by 192.168.x.x or 127.0.0.1 ? > 4- ipaddress i will write it in > > storm.zookeeper.servers: > - "ipaddress" > nimbus.host: "ipaddress" > drpc.servers: > - "localhost" > > are the same ? > 5- what do you mean by host.name ? > 6- in file /etc/hosts > 127.0.0.1 localhost > 127.0.1.1 jhk > > # The following lines are desirable for IPv6 capable hosts > #::1 ip6-localhost ip6-loopback > #fe00::0 ip6-localnet > #ff00::0 ip6-mcastprefix > #ff02::1 ip6-allnodes > #ff02::2 ip6-allrouters > should i hashing ipv6 ? > > On Thu, Dec 24, 2015 at 9:17 PM, Harsha <[email protected]> wrote: > >> 1- is production. mode means distributed mode ? >> Yes. Typically more than one machine. But I am suggesting if you want >> to test this topology you can do so by using single node cluster as >> well. >> 2. Typical production clusters usually have more than 1 node. >> 3. Usually there will be one config file that gets deployed across the >> nodes and on each node you choose start a nimbus and other nodes will >> have supervisors. >> 4. The config you've is suffice and defaults will takes place for the >> configs you don't mention. >> 5. drpc.servers: >> - "localhost" you should replace this host.name >> >> >> >> -Harsha >> >> On Thu, Dec 24, 2015, at 11:12 AM, researcher cs wrote: >> > thanks i checked the link it's good for me but the question that i'm >> > searching about >> > 1- is production mode means distributed mode ? >> > 2- if question 1 = yes , is that mean working on more than one machine ? >> > 3- if question 2 = yes , is there specific configuration should i wrote >> > in >> > storm.yaml and zoo.cfg means is that configuration right >> > storm.zookeeper.servers: >> > - "ipaddress" >> > nimbus.host: "ipaddress" >> > storm.zookeeper.port: 2181 >> > >> > storm.local.dir: /home/storm >> > >> > supervisor.slots.ports: >> > - 6700 >> > - 6701 >> > - 6702 >> > - 6703 >> > drpc.port: 3772 >> > drpc.servers: >> > - "localhost" >> > >> > and zoo.cfg is >> > tickTime=2000 >> > dataDir=/home/storm/zookeeper >> > clientPort=2181 >> > initLimit=5 >> > syncLimit=2 >> > >> > sorry for lot questions i'm searching to find a good solution but i hope >> > you can help me , Thanks >> > >> > On Thu, Dec 24, 2015 at 9:06 PM, Harsha <[email protected]> wrote: >> > >> > > You can actually use single node cluster for testing this out. The >> link >> > > I gave you have instructions to start all the required storm services >> > > and once thats up and running you can use storm jar to deploy the >> > > topology. >> > > >> > > On Thu, Dec 24, 2015, at 10:58 AM, researcher cs wrote: >> > > > thanks for replying , but what i mean that i found project using >> storm >> > > > and >> > > > coder wrote this lines >> > > > >> > > > *To compile and run the project in local mode*, type the following >> > > > commands >> > > > while being on the project root directory >> > > > mvn package >> > > > mvn compile exec:java -Dexec.classpathScope=compile >> > > > -Dexec.mainClass=trident.topology_name >> > > > >> > > > *To run in production cluster*, first package the code into a jar by >> > > > running >> > > > mvn package. >> > > > This will package your code into a jar at the path >> > > > target/Topology_name-{version}-jar-with-dependencies.jar. >> > > > Then you can submit your jar to the cluster using the storm client: >> > > > storm jar >> target/Topology_namr-1.0-SNAPSHOT-jar-with-dependencies.jar >> > > > trident.Topology_name >> > > > >> > > > Please note that in the production cluster mode, you require a >> DRPCClient >> > > > to feed the topology with tweets and get results. >> > > > >> > > > >> > > > i ran it in local mode by the above commands and worked well but >> now i >> > > > want >> > > > to run it in distrbuted mode as he wote i need to submit the >> topology to >> > > > storm but what i should need in configuration for zoo.cfg and >> storm.yaml >> > > > for working ? >> > > > >> > > > should i need more than single machine or not ? >> > > > >> > > > how it can be in distributed mode in single machine ? >> > > > >> > > > >> > > > Thanks for your help >> > > > >> > > > >> > > > >> > > > On Thu, Dec 24, 2015 at 7:43 PM, Harsha <[email protected]> wrote: >> > > > >> > > > > In storm, local mode means you run by using LocalCluster class >> its a >> > > > > simulated cluster for testing topologies and aid development of >> > > > > topologies . Example of LocalCluster >> > > > > >> > > > > >> > > >> https://github.com/apache/storm/blob/master/examples/storm-starter/src/jvm/storm/starter/WordCountTopology.java#L99 >> > > > > >> > > > > In production it would be a distributed cluster. One needs to >> setup a >> > > > > distributed cluster and config you showed seems ok . >> > > > > You can follow steps here setup up a single node cluster >> > > > > >> http://blog.harsha.io/setting-up-a-single-node-apache-storm-cluster/ >> > > > > and same can be extended for multi-node cluster. >> > > > > >> > > > > I am not quite sure about what you mean by importing a topology >> that >> > > run >> > > > > in two modes in local or production. In the first link I gave you >> , >> > > > > wordcount topology can run local cluster and on distributed as >> well. >> > > > > >> > > > > -Harsha >> > > > > >> > > > > On Thu, Dec 24, 2015, at 08:54 AM, researcher cs wrote: >> > > > > > can i find help ? >> > > > > > >> > > > > > On Thu, Dec 24, 2015 at 7:08 AM, researcher cs >> > > > > > <[email protected]> >> > > > > > wrote: >> > > > > > >> > > > > > > I want to import a project that run in two modes "local and >> > > production" >> > > > > > > mode >> > > > > > > >> > > > > > > want to get what is mean by production mode is that mean in >> > > cluster or >> > > > > > > distributed mode ? >> > > > > > > >> > > > > > > and if that right . are the configurations in storm.yaml is >> like >> > > that >> > > > > or >> > > > > > > not ? >> > > > > > > storm.zookeeper.servers: >> > > > > > > - "ipaddress" >> > > > > > > nimbus.host: "ipaddress" >> > > > > > > storm.zookeeper.port: 2181 >> > > > > > > >> > > > > > > storm.local.dir: /home/storm >> > > > > > > >> > > > > > > supervisor.slots.ports: >> > > > > > > - 6700 >> > > > > > > - 6701 >> > > > > > > - 6702 >> > > > > > > - 6703 >> > > > > > > drpc.port: 3772 >> > > > > > > drpc.servers: >> > > > > > > - "localhost" >> > > > > > > >> > > > > > > and zoo.cfg is >> > > > > > > tickTime=2000 >> > > > > > > dataDir=/home/storm/zookeeper >> > > > > > > clientPort=2181 >> > > > > > > initLimit=5 >> > > > > > > syncLimit=2 >> > > > > > > >> > > > > >> > > >> > >
