Welcome to the new Gigaspaces XAP forum. To recover your account, please follow these instructions.

Ask Your Question
0

Problem in the mirror service: NoSuchReplicationGroupExistException

Hi,

I was trying to deploy an application on Gigaspaces XAP 8.0.6 premium but I'm getting an exception when I start the agent without a global.lus and global.gsm. The very simplest case, when I try to deploy only a mirror service and a space. If I start the GSA with

./gs-agent.sh gsa.gsc 2 gsa.global.lus 1 gsa.global.gsm 1

it works properly, but when I try to start the agent with

./gs-agent.sh gsa.gsc 2 gsa.global.lus 0 gsa.global.gsm 0 gsa.gsm 1 gsa.lus 1

and then deploy the mirror and the space, I'm getting this exception:

Replication Channel: error occurred while performing handshake, replication is disabled until the error is resolved: [ source: pppDataSpace_container2:pppDataSpace ] target: data-mirror-service_container:data-mirror-service ; target url: jini://*/data-mirror-service_container/data-mirror-service?mirror=true&schema=persistent&id=2&total_members=2,1&cluster_schema=partitioned-sync2backup&locators=payuapp01:7102&groups=payuapp01-714&state=started&timeout=5000
[ target machine connection url: NIO://10.50.100.171:47627/pid[25482]/2072921898154139_3_2985078810779834278 ] [ replication group name: primary-backup-reliable-async-mirror-2 ]; Caused by: com.gigaspaces.internal.cluster.node.impl.groups.NoSuchReplicationGroupExistException: There is no replication target group under the name primary-backup-reliable-async-mirror-2. Registered groups = [primary-backup-reliable-async-mirror-1]. at com.gigaspaces.internal.cluster.node.impl.groups.ReplicationNodeGroupsHolder.getTargetGroup(ReplicationNodeGroupsHolder.java:154) at com.gigaspaces.internal.cluster.node.impl.ReplicationNode.getReplicationTargetGroup(ReplicationNode.java:709) at com.gigaspaces.internal.cluster.node.impl.packets.ConnectChannelPacket.accept(ConnectChannelPacket.java:53) at com.gigaspaces.internal.cluster.node.impl.packets.ConnectChannelPacket.accept(ConnectChannelPacket.java:29) at com.gigaspaces.internal.cluster.node.impl.ReplicationNode.onReplication(ReplicationNode.java:605) at com.gigaspaces.internal.cluster.node.impl.router.AbstractConnectionProxyBasedReplicationRouter$ConnectionEndpoint.dispatch(AbstractConnectionProxyBasedReplicationRouter.java:369) at com.gigaspaces.internal.cluster.node.impl.router.spacefinder.IReplicationConnectionProxyGigaspacesMethodinternalInvoke0.internalInvoke(Unknown Source) at com.gigaspaces.internal.reflection.fast.AbstractMethod.invoke(AbstractMethod.java:41) at com.gigaspaces.lrmi.LRMIRuntime.invoked(LRMIRuntime.java:398) at com.gigaspaces.lrmi.nio.Pivot.consumeAndHandleRequest(Pivot.java:458) at com.gigaspaces.lrmi.nio.Pivot.handleRequest(Pivot.java:552) at com.gigaspaces.lrmi.nio.Pivot$ChannelEntryTask.run(Pivot.java:179) at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:441) at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303) at java.util.concurrent.FutureTask.run(FutureTask.java:138) 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:662)

To deploy the space, the schema configuration is:
schema=partitioned-sync2backup total_members=1,1

The system is configured for Unicast discovery, the lookup groups and locators are setted properly, no replication yet (only 1 host).

My aim is to configure the system for replication with 2 host, with an schema of 2,1, but in the docs says that I shouldn't use a global lus but a normal lus (only in one host).

Why could that exception be happening?.

Thanks!.

Jean Paul Manjarres. h4. Attachments

[mirror-pu.xml|/upfiles/13759701825837982.xml]

[space-pu.xml|/upfiles/13759701821830085.xml]

{quote}This thread was imported from the previous forum. For your reference, the original is [available here|http://forum.openspaces.org/thread.jspa?threadID=4065]{quote}

asked 2013-04-19 18:29:48 -0600

Paul Manjarres gravatar image

updated 2013-08-08 09:52:00 -0600

jaissefsfex gravatar image
edit retag flag offensive close merge delete

2 Answers

Sort by ยป oldest newest most voted
0

Are you sure you set the LOOKUPLOCATORS to have the IP of the machine running the LUS before you start the agent?

Attachments

  1. mirror-pu.xml
  2. space-pu.xml

answered 2013-04-25 08:15:01 -0600

shay hassidim gravatar image
edit flag offensive delete link more

Comments

Hi, thanks for the answer.

The LOOKUPLOCATORS are setted correctly, it seems that our problem was that the mirror services were been started without cluster params, e.g. no backup. We changed the command to

gs.sh pudeploy -cluster schema=partitioned-sync2backup total_members=1,1 app-mirror-service

and now it works.

Thanks.

Attachments

  1. mirror-pu.xml
  2. space-pu.xml
Paul Manjarres gravatar imagePaul Manjarres ( 2013-05-03 17:23:31 -0600 )edit

Hi,

The pudeploy command you're using provisions two mirror service instances which is wrong. You should only have one instance.
The proper way to deploy the mirror pu would be to use the pudeploy command without the schema you have specified and have the correct mirror settings in the mirror's pu.xml file.

It would be helpful if you could post your mirror's pu.xml file.

Idan

Attachments

  1. mirror-pu.xml
  2. space-pu.xml
idan gravatar imageidan ( 2013-05-06 02:39:59 -0600 )edit

Hi, thanks for your answer.

I'm attaching the PU.xml for the space and the mirror. (I ommited irrevelant parameters) My understanding is that the mirror service should not be clustered, just to have 1 instance but we were trying to set a backup in case of failure with the "gs.sh pudeploy" command.

Also, we are getting a lot of warnings and alerts in the Web Console (gs-web.sh), "Replication Redo Log Size Alert","Replication Redo Log Overflow to Disk Alert", don't know if it is related.

Thanks in advace.

Jean Paul Manjarres Correal. h4. Attachments

[mirror-pu.xml|/upfiles/13759701838519293.xml]

[space-pu.xml|/upfiles/13759701834495493.xml]

Paul Manjarres gravatar imagePaul Manjarres ( 2013-05-06 16:42:11 -0600 )edit

Hi,

Its not necessary to have a backup for the mirror processing unit since it is stateless. The PU will be redeployed automatically on failure.

Are you also seeing the alerts you have mentioned after correctly deploying the mirror processing unit? You can find more information about these alerts here: [ http://wiki.gigaspaces.com/wiki/displ... ]

FYI - Since GigaSpaces version 9.5, there's a new space persistency API which replaces the external data source API. You can read about it here: [ http://wiki.gigaspaces.com/wiki/displ... ]

Space configuration example: [ http://wiki.gigaspaces.com/wiki/displ... ] Mirror configuration example: [ http://wiki.gigaspaces.com/wiki/displ... ]

Idan

idan gravatar imageidan ( 2013-05-07 03:33:18 -0600 )edit
0

Hi,

The exception is probably thrown because of a wrong configuration. Make sure you configure your mirror service as described here: [ http://wiki.gigaspaces.com/wiki/displ... ]

Also make sure your mirror pu's sla is configured to have a single instance.

If this doesn't help, please post your pu.xml files.

Idan

answered 2013-04-21 07:46:02 -0600

idan gravatar image
edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools

1 follower

Stats

Asked: 2013-04-19 18:29:48 -0600

Seen: 914 times

Last updated: Apr 25 '13