[
https://issues.apache.org/jira/browse/FLINK-24031?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17567043#comment-17567043
]
Yang Wang commented on FLINK-24031:
-----------------------------------
>From your attached logs, the JobManager is listening at correct address. And
>the TaskManager tries to register to JM with "flink-jobmanager". However, I
>find an unknown host exception in TM logs. So I think the root cause might be
>coreDNS. You need to verify you coreDNS works normally. And you could start a
>busybox and do a nslookup for verification.
{code:java}
Caused by: [java.net.UnknownHostException: flink-jobmanager: Name or service
not known {code}
If you could not manage the coreDNS works, then using a Flink Kubernetes HA is
another option. Because the TaskManager will retrieve the JobManager leader
address via HA service, not the K8s service. And then register to the
JobManager using pod ip.
[https://nightlies.apache.org/flink/flink-docs-release-1.14/docs/deployment/resource-providers/standalone/kubernetes/#high-availability-with-standalone-kubernetes]
jobmanager-application-ha.yaml
> I am trying to deploy Flink in kubernetes but when I launch the taskManager
> in other container I get a Exception
> ----------------------------------------------------------------------------------------------------------------
>
> Key: FLINK-24031
> URL: https://issues.apache.org/jira/browse/FLINK-24031
> Project: Flink
> Issue Type: Bug
> Components: Deployment / Kubernetes
> Affects Versions: 1.13.0, 1.13.2
> Reporter: Julio Pérez
> Priority: Major
> Labels: pull-request-available
> Attachments: JM.log, TM.log, flink-map.yml, jobmanager.log,
> jobmanager.yml, taskmanager.log, taskmanager.yml
>
>
> I explain here -> [https://github.com/apache/flink/pull/17020]
> I have a problem when I try to run Flink in k8s with the follow manifests
> I have the following exception
> # JobManager :
> {quote}2021-08-27 09:16:57,917 ERROR akka.remote.EndpointWriter [] - dropping
> message [class akka.actor.ActorSelectionMessage] for non-local recipient
> [Actor[akka.tcp://flink@jobmanager-hs:6123/]] arriving at
> [akka.tcp://flink@jobmanager-hs:6123] inbound addresses are
> [akka.tcp://flink@cluster:6123]
> 2021-08-27 09:17:01,255 DEBUG
> org.apache.flink.runtime.resourcemanager.StandaloneResourceManager [] -
> Trigger heartbeat request.
> 2021-08-27 09:17:01,284 DEBUG
> org.apache.flink.runtime.resourcemanager.StandaloneResourceManager [] -
> Trigger heartbeat request.
> 2021-08-27 09:17:10,008 DEBUG akka.remote.transport.netty.NettyTransport []
> - Remote connection to [/172.17.0.1:34827] was disconnected because of [id:
> 0x13ae1d03, /172.17.0.1:34827 :> /172.17.0.23:6123] DISCONNECTED
> 2021-08-27 09:17:10,008 DEBUG akka.remote.transport.ProtocolStateActor [] -
> Association between local [tcp://flink@cluster:6123] and remote
> [tcp://[email protected]:34827] was disassociated because the
> ProtocolStateActor failed: Unknown
> 2021-08-27 09:17:10,009 WARN akka.remote.ReliableDeliverySupervisor [] -
> Association with remote system [akka.tcp://[email protected]:6122] has
> failed, address is now gated for [50] ms. Reason: [Disassociated]
> {quote}
> TaskManager:
> {quote}INFO org.apache.flink.runtime.taskexecutor.TaskExecutor [] - Could not
> resolve ResourceManager address
> akka.tcp://flink@flink-jobmanager:6123/user/rpc/resourcemanager__, retrying
> in 10000 ms: Could not connect to rpc endpoint under address
> akka.tcp://flink@flink-jobmanager:6123/user/rpc/resourcemanager__.
> INFO org.apache.flink.runtime.taskexecutor.TaskExecutor [] - Could not
> resolve ResourceManager address
> akka.tcp://flink@flink-jobmanager:6123/user/rpc/resourcemanager__, retrying
> in 10000 ms: Could not connect to rpc endpoint under address
> akka.tcp://flink@flink-jobmanager:6123/user/rpc/resourcemanager__.
> {quote}
> Best regards,
> Julio
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
