Confluent Control Center - managing multiple clusters

[singh.sh...@gmail.com]

We have a requirement to have a single Control Center manage multiple clusters ..

is that possible ? what are the configuration steps for this ?

[Gwen Shapira]

It is possible.

Basically, configuring control center for one cluster involves:

1. Installing control center
2. Installing the monitoring plugin on every broker you want to monitor
3. Installing interceptors on every client you want to monitor

If you want to monitor multiple clusters, you simply repeat step 2 on all the clusters you want to monitor.

Gwen

On Thu, Sep 14, 2017 at 11:39 AM <singh.sh...@gmail.com> wrote:

We have a requirement to have a single Control Center manage multiple clusters ..

is that possible ? what are the configuration steps for this ?

--
You received this message because you are subscribed to the Google Groups "Confluent Platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to confluent-platf...@googlegroups.com.
To post to this group, send email to confluent...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/confluent-platform/9a8236e6-0fc7-4756-a2e2-7c7c13bde9f3%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[singh.sh...@gmail.com]

Hello Gwen - 

thanks for the response .. i've made the changes you listed out, and am able to monitor the Brokers in their respective control-centers.

However, the scenario of managing/monitoring multiple clusters using a single Control Center is not yet working...

I've the Control Center running on Cluster1 

& i'm trying to enable monitoring of Cluster2 brokers using the Control Center (on Cluster1).  

Changes made --

On Cluster2, i've made the following changes in server.properties file (i.e pointed the Brokers to Metrics Reporter on Cluster1) ..

metric.reporter=io.confluent.metrics.reporter.ConfluentMetricsReporter 

confluent.metrics.reporter.bootstrap.servers=<cluster1-broker>:<port no>

 

 

confluent.metrics.reporter.zookeeper.connect=<cluster1-zookeeper>:<port no>

I don't see the ClusterId2 in the "Clusters" page on Control Center UI..

Do i need to make any other changes ?

(for eg. in the control-center.properties file on Cluster1)    

Also, If there is documentation around this, pls point me to the link.. thnx.

On Thursday, September 14, 2017 at 11:42:09 AM UTC-7, Gwen Shapira wrote:

It is possible.

Basically, configuring control center for one cluster involves:

1. Installing control center
2. Installing the monitoring plugin on every broker you want to monitor
3. Installing interceptors on every client you want to monitor

If you want to monitor multiple clusters, you simply repeat step 2 on all the clusters you want to monitor.

Gwen

On Thu, Sep 14, 2017 at 11:39 AM <singh.sh...@gmail.com> wrote:

We have a requirement to have a single Control Center manage multiple clusters ..

is that possible ? what are the configuration steps for this ?

--
You received this message because you are subscribed to the Google Groups "Confluent Platform" group.

To unsubscribe from this group and stop receiving emails from it, send an email to confluent-platform+unsub...@googlegroups.com.

[singh.sh...@gmail.com]

hi Gwen, experts -

here is the update on this ..

I was able to setup the Confluent Control Center to mange multiple kafka clusters, using the below mentioned suggested settings in broker server.properties..   

metric.reporter=io.confluent.metrics.reporter.ConfluentMetricsReporter 

confluent.metrics.reporter.bootstrap.servers=<cluster1-broker>:<port no>

confluent.metrics.reporter.zookeeper.connect=<cluster1-zookeeper>:<port no> 

confluent.metrics.reporter.zookeeper.connect=<cluster1-zookeeper>:<port no>

However, i did notice couple of issues ..

1)

I've cleared up zookeeper (in trying to resolve zookeeper related issues), and therefore the clusterId has changed for the cluster. On restarting the control center & other services, Control Center seems to retain the older clusterIds... pls see the screen attached.

in the screenshot, of the 5 clusterIds - only 3 of them actually exist currently .

Any ideas on how to reset the cache ? i tried restarting the Control Center, but have not had much success ..

2) On the Control Center Cluster page, for the remote cluster, the "zookeeper servers" & "bootstrap servers" are shown to be None. (screenshot attached)

Is that expected ?

I think these correspond to the entries in the control-center.properties file .. (in which case, it makes sense )

Pls. confirm my understanding.