[jira] [Commented] (NIFI-9695) Load balancing not working
[
https://issues.apache.org/jira/browse/NIFI-9695?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17542366#comment-17542366
]
Jake Dalli commented on NIFI-9695:
--
[~joewitt] ok thank you, in the mean time is it possible to understand the
issue and what the impact is?
> Load balancing not working
> --
>
> Key: NIFI-9695
> URL: https://issues.apache.org/jira/browse/NIFI-9695
> Project: Apache NiFi
> Issue Type: Bug
>Affects Versions: 1.13.2
>Reporter: davide
>Priority: Major
> Labels: Cluster, balancer, nifi
> Attachments: activeThread.png, cores.png, flow.png, heapUse.png,
> version.png
>
>
> I have a 3 nodes cluster(as you can see on attached images) with a connection
> with a Load Balance Strategy set to Partition by attribute, sometimes some
> flowfiles get stuck in this connection (see image1). Furthermore we have this
> log error in bulletin board
> {\{10:56:28 UTC ERROR -xxx-xx-nifi0.:8443 }}
> {{Failed to receive FlowFiles for Load Balancing due to
> org.apache.nifi.controller.queue.clustered.server.TransactionAbortedException:
> Attempted to receive FlowFiles from Peer -xxx-xx-nifi2. for
> Connection with ID 0c20343f-f63d-1ef1-8da0-6d3cb9fc670d but no Connection
> exists with that ID}}
> {{10:56:28 UTC ERROR -xxx-xx-nifi0.:8443}}
> {{{}Failed to communicate with Peer -xxx-xx-nifi2.:8443 when load
> balancing data for Connection with ID c12f3c02-9f05-1232--f8d925a9
> due to java.io.IOException: Failed to decrypt data from Peer
> -xxx-xx-nifi2.:8443 because Peer unexpectedly closed
> connection{}}}}}{}}}
>
> The weird thing is that only some flowfiles are stucked other flowfiles goes
> forward in the flows.
> I think this bug is the same of
> https://issues.apache.org/jira/browse/NIFI-9574, I post a comment in
> NIFI-9574 but I have got no answer.
>
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
[jira] [Commented] (NIFI-9695) Load balancing not working
[
https://issues.apache.org/jira/browse/NIFI-9695?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17542362#comment-17542362
]
Jake Dalli commented on NIFI-9695:
--
We are also experiencing this issue. [~vadidde] [~joewitt] how does this impact
the service and do you know any quick fixes?
> Load balancing not working
> --
>
> Key: NIFI-9695
> URL: https://issues.apache.org/jira/browse/NIFI-9695
> Project: Apache NiFi
> Issue Type: Bug
>Affects Versions: 1.13.2
>Reporter: davide
>Priority: Major
> Labels: Cluster, balancer, nifi
> Attachments: activeThread.png, cores.png, flow.png, heapUse.png,
> version.png
>
>
> I have a 3 nodes cluster(as you can see on attached images) with a connection
> with a Load Balance Strategy set to Partition by attribute, sometimes some
> flowfiles get stuck in this connection (see image1). Furthermore we have this
> log error in bulletin board
> {\{10:56:28 UTC ERROR -xxx-xx-nifi0.:8443 }}
> {{Failed to receive FlowFiles for Load Balancing due to
> org.apache.nifi.controller.queue.clustered.server.TransactionAbortedException:
> Attempted to receive FlowFiles from Peer -xxx-xx-nifi2. for
> Connection with ID 0c20343f-f63d-1ef1-8da0-6d3cb9fc670d but no Connection
> exists with that ID}}
> {{10:56:28 UTC ERROR -xxx-xx-nifi0.:8443}}
> {{{}Failed to communicate with Peer -xxx-xx-nifi2.:8443 when load
> balancing data for Connection with ID c12f3c02-9f05-1232--f8d925a9
> due to java.io.IOException: Failed to decrypt data from Peer
> -xxx-xx-nifi2.:8443 because Peer unexpectedly closed
> connection{}}}}}{}}}
>
> The weird thing is that only some flowfiles are stucked other flowfiles goes
> forward in the flows.
> I think this bug is the same of
> https://issues.apache.org/jira/browse/NIFI-9574, I post a comment in
> NIFI-9574 but I have got no answer.
>
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
[jira] [Created] (NIFI-9930) Add AWS Glue Schema Registry to Nifi Record Utils
Jake Dalli created NIFI-9930: Summary: Add AWS Glue Schema Registry to Nifi Record Utils Key: NIFI-9930 URL: https://issues.apache.org/jira/browse/NIFI-9930 Project: Apache NiFi Issue Type: Wish Components: Core Framework Reporter: Jake Dalli Currently NiFi supports Confluent Schema Registry and Hortonworks Schema Registry.Many NiFi Users run NiFi inside AWS, which has its own schema registry via AWS Glue: AWS Glue Schema Registry. As a developer, I would like to use AWS Glue Schema Registry within NiFi. AWS Documentation: https://docs.aws.amazon.com/glue/latest/dg/schema-registry.html -- This message was sent by Atlassian Jira (v8.20.1#820001)
[jira] [Created] (NIFI-7915) Components using okhttp 3.x Broken on Nifi 1.9.2
Jake Dalli created NIFI-7915:
Summary: Components using okhttp 3.x Broken on Nifi 1.9.2
Key: NIFI-7915
URL: https://issues.apache.org/jira/browse/NIFI-7915
Project: Apache NiFi
Issue Type: Bug
Components: Core Framework
Affects Versions: 1.9.2
Reporter: Jake Dalli
Hi,
As of the latest Java version (8u252), components using the okhttp3 dependency
are broken with the following error:
{code:java}
clientBuilder.sslSocketFactory(SSLSocketFactory) not supported on JDK 9
{code}
Following some investigation, it turns out to be an issue with the dependency:
[https://github.com/square/okhttp/issues/6019]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
[jira] [Created] (NIFI-6701) PublishGCPPubSub does not send all messages
Jake Dalli created NIFI-6701:
Summary: PublishGCPPubSub does not send all messages
Key: NIFI-6701
URL: https://issues.apache.org/jira/browse/NIFI-6701
Project: Apache NiFi
Issue Type: Bug
Components: Extensions
Affects Versions: 1.9.2
Reporter: Jake Dalli
I have found an issue with the PublishGCPPubSub processor. After implementing
a number of flows, I have realised that approximately 30% of messages are
sporadically not sent to Pubsub in spite of a 'Send Successful' response from
the processor.
Upon investigating the code on github
([https://github.com/apache/nifi/blob/master/nifi-nar-bundles/nifi-gcp-bundle/nifi-gcp-processors/src/main/java/org/apache/nifi/processors/gcp/pubsub/PublishGCPubSub.java|https://slack-redir.net/link?url=https%3A%2F%2Fgithub.com%2Fapache%2Fnifi%2Fblob%2Fmaster%2Fnifi-nar-bundles%2Fnifi-gcp-bundle%2Fnifi-gcp-processors%2Fsrc%2Fmain%2Fjava%2Forg%2Fapache%2Fnifi%2Fprocessors%2Fgcp%2Fpubsub%2FPublishGCPubSub.java]),
I found the following issue:
# On line 163 we get the Future for the 'Publish message' action.
# On line 165 we have the following loop:
{code:java}
while (messageIdFuture.isDone()) {
Thread.sleep(500L);
}
{code}
This means that if the future isDone method returns, the thread sleeps. It is
unclear to me why this is the case and I believe this is a bag, in my opinion
the processor should sleep if the future is not done, thus the loop should read:
{code:java}
while(!messageIdFuture.isDone()){ ... }{code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
