Discovery module 0.8.0

I am working with ES 0.8.0.
I have cluster with two nodes

Node1 – ES server - data not client
Node2 - Java application with embedded ES node – client, no data

Discovery mechanism working fine (master switch and nodes restart) when I configure both nodes to work with JGROUP UDP discovery.

When I try to work with ZEN or JGROUP TCP “master” switch does not work.

My scenario was:

Start Node1 first.
Now I can start/stop Node2 and it will connect to master as well as first time each time and no any problem.
I stopped Node1, Node2 got “Master”.
Start again Node1 then:

Jgroup TCP discovery:
Node1 was started as master without connection to Node2 . I got two masters.

Zen discovery:
I got the following error

[22:04:31,679][INFO ][node ] [Everyman] {ElasticSearch/0.8.0}[6264]: Initializing ...
[22:04:31,688][INFO ][plugins ] [Everyman] Loaded [discovery-jgroups]
[22:04:32,292][INFO ][node ] [Everyman] {ElasticSearch/0.8.0}[6264]: Initialized
[22:04:32,292][INFO ][node ] [Everyman] {ElasticSearch/0.8.0}[6264]: Starting ...
[22:04:32,457][INFO ][transport ] [Everyman] bound_address[inet[/0.0.0.0:9300]], publish_address[inet[/192.168.0.175:9300]]
[22:04:35,611][WARN ][discovery.zen ] [Everyman] Failed to send join request to master [[Smythe, Alistair][9999fa3d-c91b-4426-97a6-3a067d5fd51f][inet[/192.168.0.175:9300]]], retrying...
org.elasticsearch.transport.RemoteTransportException: [Everyman][inet[/192.168.0.175:9300]][discovery/zen/join]
Caused by: org.elasticsearch.ElasticSearchIllegalStateException: Node [[Everyman][7f245178-8d30-4448-9dc9-8092a7c01372][inet[/192.168.0.175:9300]]] not master for join request from [[Everyman][7f245178-8d30-4
:9300]]]
at org.elasticsearch.discovery.zen.ZenDiscovery.handleJoinRequest(ZenDiscovery.java:386)
at org.elasticsearch.discovery.zen.ZenDiscovery.access$900(ZenDiscovery.java:58)
at org.elasticsearch.discovery.zen.ZenDiscovery$MembershipListener.onJoin(ZenDiscovery.java:450)
at org.elasticsearch.discovery.zen.membership.MembershipAction$JoinRequestRequestHandler.messageReceived(MembershipAction.java:113)
at org.elasticsearch.discovery.zen.membership.MembershipAction$JoinRequestRequestHandler.messageReceived(MembershipAction.java:104)
at org.elasticsearch.transport.netty.MessageChannelHandler$3.run(MessageChannelHandler.java:175)
at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
at java.lang.Thread.run(Thread.java:619)
[22:04:38,631][WARN ][discovery.zen ] [Everyman] Failed to send join request to master [[Smythe, Alistair][9999fa3d-c91b-4426-97a6-3a067d5fd51f][inet[/192.168.0.175:9300]]], retrying...
org.elasticsearch.transport.RemoteTransportException: [Everyman][inet[/192.168.0.175:9300]][discovery/zen/join]
Caused by: org.elasticsearch.ElasticSearchIllegalStateException: Node [[Everyman][7f245178-8d30-4448-9dc9-8092a7c01372][inet[/192.168.0.175:9300]]] not master for join request from [[Everyman][7f245178-8d30-4
:9300]]]

My configuration file
cluster:
name:

transport:
netty:
reuse_address: false

discovery:
type: zen or jgroups
jgroups:
config: tcp
bind_port: 7800
bind_addr:
tcpping:
initial_hosts: [7800]

Client node configuration

	// General settings
	Properties props = System.getProperties();
	props.setProperty("java.net.preferIPv4Stack", "true");
	props.setProperty("transport.netty.reuse_address", "false");
	System.setProperties(props);

	//Create settings builder
	ImmutableSettings.Builder builder = ImmutableSettings.settingsBuilder();

	if (discoveryType.equals(DISCOVERY_JGROUP_TCP)){
		logger.info("Jgroups " + DISCOVERY_JGROUP_TCP + " discovery will start");
		builder.put("discovery.type", "jgroups");
		builder.put("discovery.jgroups.config", "tcp");
		builder.put("discovery.jgroups.bind_port", JGROUP_TCP_PORT);
		builder.put("discovery.jgroups.bind_addr", hostName);
		builder.put("discovery.jgroups.tcpping.initial_hosts", hostName	+ "["+ JGROUP_TCP_PORT +"]");

	}else if (discoveryType.equals(DISCOVERY_JGROUP_UDP)){
		logger.info("Jgroups " + DISCOVERY_JGROUP_UDP + "discovery will start");
		builder.put("discovery.type", "jgroups");
		builder.put("discovery.jgroups.config", "udp");

	}else{
		//Set discovery to DISCOVERY_ZEN default
		logger.info("Default " + DISCOVERY_ZEN + "discovery will start");
		builder.put("discovery.type", "zen");
	}

I try to get cluster nodes status with
curl -XGET 'http://localhost:9200/_cluster/nodes'
So, when only Node1 running I got node status . When Node2 was connected to cluster I got
repose like
{"error":"JsonGenerationException[Can not write a field name, expecting a value]"}

Thanks
Lena

I don't see where you configure for zen discovery the lists of hosts it uses
(I assume you don't want to use multicast)? For example:

discovery.zen.ping.unicast.hosts: host1:9300,host2:9300,host3:9300

When you close the nodes, is that through graceful shutdown (kill pid, or
Ctrl+c), or through kill -9?

-shay.banon

On Wed, Jun 2, 2010 at 10:33 PM, Lenak lenakovganov@gmail.com wrote:

I am working with ES 0.8.0.
I have cluster with two nodes

Node1 – ES server - data not client
Node2 - Java application with embedded ES node – client, no data

Discovery mechanism working fine (master switch and nodes restart) when I
configure both nodes to work with JGROUP UDP discovery.

When I try to work with ZEN or JGROUP TCP “master” switch does not work.

My scenario was:

Start Node1 first.
Now I can start/stop Node2 and it will connect to master as well as first
time each time and no any problem.
I stopped Node1, Node2 got “Master”.
Start again Node1 then:

Jgroup TCP discovery:
Node1 was started as master without connection to Node2 . I got two
masters.

Zen discovery:
I got the following error

[22:04:31,679][INFO ][node ] [Everyman]
{Elasticsearch/0.8.0}[6264]: Initializing ...
[22:04:31,688][INFO ][plugins ] [Everyman] Loaded
[discovery-jgroups]
[22:04:32,292][INFO ][node ] [Everyman]
{Elasticsearch/0.8.0}[6264]: Initialized
[22:04:32,292][INFO ][node ] [Everyman]
{Elasticsearch/0.8.0}[6264]: Starting ...
[22:04:32,457][INFO ][transport ] [Everyman]
bound_address[inet[/0.0.0.0:9300]],
publish_address[inet[/192.168.0.175:9300]]
[22:04:35,611][WARN ][discovery.zen ] [Everyman] Failed to send
join request to master [[Smythe,
Alistair][9999fa3d-c91b-4426-97a6-3a067d5fd51f][inet[/192.168.0.175:9300
]]],
retrying...
org.elasticsearch.transport.RemoteTransportException:
[Everyman][inet[/192.168.0.175:9300]][discovery/zen/join]
Caused by: org.elasticsearch.ElasticSearchIllegalStateException: Node
[[Everyman][7f245178-8d30-4448-9dc9-8092a7c01372][inet[/192.168.0.175:9300
]]]
not master for join request from [[Everyman][7f245178-8d30-4
:9300]]]
at

org.elasticsearch.discovery.zen.ZenDiscovery.handleJoinRequest(ZenDiscovery.java:386)
at

org.elasticsearch.discovery.zen.ZenDiscovery.access$900(ZenDiscovery.java:58)
at

org.elasticsearch.discovery.zen.ZenDiscovery$MembershipListener.onJoin(ZenDiscovery.java:450)
at

org.elasticsearch.discovery.zen.membership.MembershipAction$JoinRequestRequestHandler.messageReceived(MembershipAction.java:113)
at

org.elasticsearch.discovery.zen.membership.MembershipAction$JoinRequestRequestHandler.messageReceived(MembershipAction.java:104)
at

org.elasticsearch.transport.netty.MessageChannelHandler$3.run(MessageChannelHandler.java:175)
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:619)
[22:04:38,631][WARN ][discovery.zen ] [Everyman] Failed to send
join request to master [[Smythe,
Alistair][9999fa3d-c91b-4426-97a6-3a067d5fd51f][inet[/192.168.0.175:9300
]]],
retrying...
org.elasticsearch.transport.RemoteTransportException:
[Everyman][inet[/192.168.0.175:9300]][discovery/zen/join]
Caused by: org.elasticsearch.ElasticSearchIllegalStateException: Node
[[Everyman][7f245178-8d30-4448-9dc9-8092a7c01372][inet[/192.168.0.175:9300
]]]
not master for join request from [[Everyman][7f245178-8d30-4
:9300]]]

My configuration file
cluster:
name:

transport:
netty:
reuse_address: false

discovery:
type: zen or jgroups
jgroups:
config: tcp
bind_port: 7800
bind_addr:
tcpping:
initial_hosts: [7800]

Client node configuration

           // General settings
           Properties props = System.getProperties();
           props.setProperty("java.net.preferIPv4Stack", "true");
           props.setProperty("transport.netty.reuse_address", "false");
           System.setProperties(props);

           //Create settings builder
           ImmutableSettings.Builder builder =

ImmutableSettings.settingsBuilder();

           if (discoveryType.equals(DISCOVERY_JGROUP_TCP)){
                   logger.info("Jgroups " + DISCOVERY_JGROUP_TCP + "

discovery will start");
builder.put("discovery.type", "jgroups");
builder.put("discovery.jgroups.config", "tcp");
builder.put("discovery.jgroups.bind_port",
JGROUP_TCP_PORT);
builder.put("discovery.jgroups.bind_addr",
hostName);

builder.put("discovery.jgroups.tcpping.initial_hosts", hostName + "["+
JGROUP_TCP_PORT +"]");

           }else if (discoveryType.equals(DISCOVERY_JGROUP_UDP)){
                   logger.info("Jgroups " + DISCOVERY_JGROUP_UDP +

"discovery will start");
builder.put("discovery.type", "jgroups");
builder.put("discovery.jgroups.config", "udp");

           }else{
                   //Set discovery to DISCOVERY_ZEN default
                   logger.info("Default " + DISCOVERY_ZEN + "discovery

will start");
builder.put("discovery.type", "zen");
}

I try to get cluster nodes status with
curl -XGET 'http://localhost:9200/_cluster/nodes'
So, when only Node1 running I got node status . When Node2 was connected
to
cluster I got
repose like
{"error":"JsonGenerationException[Can not write a field name, expecting a
value]"}

Thanks
Lena

--
View this message in context:
http://elasticsearch-users.115913.n3.nabble.com/Discovery-module-0-8-0-tp865726p865726.html
Sent from the Elasticsearch Users mailing list archive at Nabble.com.