provided by pinot repo

the k8s yaml:

Oh cool, I'll look at that.

I have a schema with these fields

{tenant_id, uidx, date, hour, steps}

and

{tenat_id, uidx, date, hour}

is part of the

primary key

. So Source debezium Kafka connector will keep pushing updates on the above schema. So my question is: How can we maintain last view value in Pinot. In this case

steps

Are there any recommended monitoring solutions for Pinot? I see this gitbook page. I see JMX to Prometheus.

The Pinot metrics can be fetched from JMX, you could also query it.

Uber has wrote a wrapper on PinotStarter and inject a JmxQueryLayer to fetch and organize metrics, then use internal lib to publish to local metrics collector agent. This is usually how people hook up with existing monitoring system

Or if you have scrapper to fetch metrics, then you can try to expose an metrics endpoint for node_exporters to scrapper the metrics

either will work, based on how your monitoring system architecture

@User Given the extensive adoption of Prometheus, maybe consider exposing the metrics in Prom compliant format (flag enabled) so that people can easily get these metrics into Prom?

I'm building a system from scratch so I'd take any advice on monitoring systems that work well with it. Sounds like Prometheus is a good solution.

@User this is good idea, I bet there must be some MetricsReporter tooling we can use

just the metrics may still be old fashion dot format, not tag based

Can I disable a server instance S in a tenant? I want to make sure segment resource no longer allocated to S and any resource in ERROR state on S becomes OFFLINE. I tried the controller api:

/instances/{instanceName}/state

but this API does not work for me.

/instances/{instanceName}

works fine. Deleting the instance does not work either because the instance is live.

New segments will not get assigned

hey all, I am getting some warning logs on the server container in K8s

The configuration '{stream config name}' was supplied but isn't a known config.

Is it natural or am I missing something?

New blog post on using Apache Pinot to analyze GitHub Events https://medium.com/apache-pinot-developer-blog/using-apache-pinot-and-kafka-to-analyze-github-events-93cdcb57d5f7 Thanks @User for building the app and @User for the blog.

Where can I find the information about query caching strategy used by pinot and how to configure it if possible?

Pinot does not cache query results, is there a doc that suggests otherwise?

Ok... there's no doc suggesting so.

Is there some reason behind not having query result caching?

One issue is cache invalidation (data push, RT ingestion). We are in the process of discussing on adding a separate caching layer, though.

We also dint find a strong need to add it, queries were fast without caching. simple query based caching is easy to add but not sure about the effectiveness.

yes sounds fair.

2020/04/14 17:31:41.007 INFO [HelixBrokerStarter] [main] Starting Pinot broker
 2020/04/14 17:31:41.013 INFO [HelixBrokerStarter] [main] Connecting spectator Helix manager
 2020/04/14 17:31:41.353 INFO [HelixBrokerStarter] [main] Setting up broker server builder
 Apr 14, 2020 5:31:44 PM org.glassfish.grizzly.http.server.NetworkListener start
 INFO: Started listener bound to [0.0.0.0:8080]
 Apr 14, 2020 5:31:44 PM org.glassfish.grizzly.http.server.HttpServer start
 INFO: [HttpServer] Started.
 2020/04/14 17:31:48.485 INFO [HelixBrokerStarter] [main] Initializing cluster change mediator
 2020/04/14 17:31:48.541 INFO [HelixBrokerStarter] [main] Connecting participant Helix manager
 2020/04/14 17:31:48.623 WARN [ParticipantHealthReportTask] [main] ParticipantHealthReportTimerTask already stopped
 2020/04/14 17:31:49.278 WARN [ConfigAccessor] [main] No config found at /olap-data-2/CONFIGS/RESOURCE/brokerResource
 2020/04/14 17:31:49.408 WARN [CallbackHandler] [main] Callback handler received event in wrong order. Listener: org.apache.helix.messaging.handling.HelixTaskExecutor@1a89f5be, path: /olap-data-2/INSTANCES/Broker_pinot-broker-1.pinot-broker-headless.olap-data-2.svc.cluster.local_8080/MESSAGES, expected types: [CALLBACK, FINALIZE] but was INIT
 2020/04/14 17:31:49.542 INFO [HelixBrokerStarter] [main] Registering service status handler
 2020/04/14 17:31:49.557 INFO [HelixBrokerStarter] [main] Finish starting Pinot broker
 2020/04/14 17:33:58.838 INFO [HelixBrokerStarter] [Thread-1] Shutting down Pinot broker
 2020/04/14 17:33:58.838 INFO [HelixBrokerStarter] [Thread-1] Shutting down time boundary info refresh message handler
 2020/04/14 17:33:58.838 INFO [HelixBrokerStarter] [Thread-1] Disconnecting participant Helix manager
 2020/04/14 17:33:58.847 WARN [StateModel] [Thread-1] Default reset method invoked. Either because the process longer own this resource or session timedout
 2020/04/14 17:33:58.847 WARN [StateModel] [Thread-1] Default reset method invoked. Either because the process longer own this resource or session timedout
 2020/04/14 17:33:58.853 WARN [StateModel] [Thread-1] Default reset method invoked. Either because the process longer own this resource or session timedout
 2020/04/14 17:33:58.853 WARN [StateModel] [Thread-1] Default reset method invoked. Either because the process longer own this resource or session timedout
 2020/04/14 17:33:59.009 INFO [HelixBrokerStarter] [Thread-1] Stopping cluster change mediator
 2020/04/14 17:33:59.009 INFO [HelixBrokerStarter] [Thread-1] Stopping broker server builder
 2020/04/14 17:34:09.010 ERROR [DataTableHandler] [nioEventLoopGroup-2-1] Channel for server: pinot-server-2_R is now inactive, marking server down
 2020/04/14 17:34:09.010 ERROR [DataTableHandler] [nioEventLoopGroup-2-2] Channel for server: pinot-server-1_R is now inactive, marking server down
 2020/04/14 17:34:09.011 ERROR [DataTableHandler] [nioEventLoopGroup-2-1] Channel for server: pinot-server-0_R is now inactive, marking server down
 2020/04/14 17:34:09.024 INFO [HelixBrokerStarter] [Thread-1] Disconnecting spectator Helix manager
 2020/04/14 17:34:09.019 WARN [AbstractChannel] [grizzly-http-server-1] Force-closing a channel whose registration task was not accepted by an event loop: [id: 0x2fc9d452]
 java.util.concurrent.RejectedExecutionException: event executor terminated
        at io.netty.util.concurrent.SingleThreadEventExecutor.reject(SingleThreadEventExecutor.java:842) ~[pinot-all-0.3.0-SNAPSHOT-jar-with-dependencies.jar:0.3.0-SNAPSHOT-183d810494f0ac87f1ac14f85bb8e04dca3e99ec]
        at io.netty.util.concurrent.SingleThreadEventExecutor.offerTask(SingleThreadEventExecutor.java:328) ~[pinot-all-0.3.0-SNAPSHOT-jar-with-dependencies.jar:0.3.0-SNAPSHOT-183d810494f0ac87f1ac14f85bb8e04dca3e99ec]

  Not able to find the cause of this failure. Thanks!

what is the cpu/memory capacity for each container

Haven't put a limit on the containers the resource utilization was as below

NAME                 CPU(cores)   MEMORY(bytes)
pinot-broker-0       59m          1564Mi
pinot-broker-1       1207m        1475Mi
pinot-broker-2       29m          1599Mi
pinot-controller-0   4m           959Mi
pinot-controller-1   26m          987Mi
pinot-controller-2   4m           928Mi
pinot-server-0       323m         2439Mi
pinot-server-1       285m         5275Mi
pinot-server-2       185m         2239Mi
pinot-zookeeper-0    14m          377Mi