Support Forums : DataStax

DataStax Support Forums » OpsCenter

are 61620/61621 ports configurable in community version of opscenter?

(19 posts) (3 voices)

Started 1 year ago by
Latest reply from thobbs
This topic is not resolved

Tags:

No tags yet.

12 Next »

Anonymous
Unregistered

these ports are blocked in our env, how can I change them to different ports?

also does opscenter support multiple clusters?

Thanks

Posted 1 year ago #
thobbs
Administrator
They are configurable. If you add the following lines to opscenterd.conf and restart opscenterd, the change should take effect:
```
[agents]
https_port = 61621
incoming_port = 61620
```
Posted 1 year ago #
thobbs
Administrator

Sorry, obviously you would change those port numbers, but those are the default values.

Posted 1 year ago #
thobbs
Administrator

Actually, I left out an important step. You will need to update the config for each agent and restart them as well.

The config file you will need to edit is address.yaml. If the agents were installed automatically by opscenter, or if you installed the agent packages (either the rpm or deb files) manually, it will be located at /var/lib/opscenter-agent/conf/address.yaml.

Add the following line to address.yaml, replacing 61620 with your custom port:

stomp_port: 61620

This port number should match the "incoming_port" number in opscenterd.conf.

Posted 1 year ago #
Anonymous
Unregistered

thanks thobbs, let me try that.

btw does opscenter work with multiple clusters?

Posted 1 year ago #
thobbs
Administrator

OpsCenter does not yet support multiple clusters, but that's something we're planning to support soon in the Enterprise version, probably some time in the first half of 2012.

Posted 1 year ago #
Anonymous
Unregistered

thanks for reply.

I changed ports to 41620/41621. I see following errors on opscenter host.
slcdbx1035 is client which has agent manually installed and it has no errors in agent.log.

Any idea?

2011-12-20 14:46:18-0700 [] ERROR: Error collecting datacenter: No agents available to query.
2011-12-20 14:47:27-0700 [] INFO: Agent for ip slcdbx1035 is version u'2.4'
2011-12-20 14:48:18-0700 [] ERROR: Error collecting datacenter: No agents available to query.

Posted 1 year ago #
thobbs
Administrator

You need to use the IP address of the agent rather than its hostname. This is also set in address.yaml.

Posted 1 year ago #
Anonymous
Unregistered

thanks, using IP worked.

Posted 1 year ago #
Anonymous
Unregistered

hi, one more question.

how does opscenter count # of reads or writes? # of request showed in opscenter graph are much lower comparing to counts from nodetool cfstats.

Posted 1 year ago #
Anonymous
Unregistered

2nd question: can opscenter client/server share same cassandra host? I tried to do that, but client log shows error as:

INFO [StompConnection receiver] 2011-12-20 18:29:14,732 Jetty server started
ERROR [Jetty] 2011-12-20 18:29:15,372 Exception running Jetty, restarting: java.net.BindException: Address already in use

Posted 1 year ago #
Anonymous
Unregistered

please ignore my 2nd question. an opscenter client/server share same cassandra host?

I forgot to kill previous client process.

Posted 1 year ago #
nickmbailey
Key Master

"how does opscenter count # of reads or writes?"

Nodetool cfstats displays the total read/write counts for column families. So these numbers are the total number of reads/writes that have been performed on a column family since the cassandra process started. OpsCenter uses this value to display the number of reads/writes per second happening in the cluster. This number more accurately represents the current state of the cluster.

"can opscenter client/server share same cassandra host?"

Yes, the agent and daemon can run on the same host. Are you using different ports for the two port settings? That error either indicates you are attempting to use the same port for both settings or something else on your machine is already listening on the agent api port. Is it possible you already have a running agent process? If you attempted to start two agent processes you would see that error.

Posted 1 year ago #
Anonymous
Unregistered

Hi Nick,

"Nodetool cfstats displays the total read/write counts for column families." at node level, not at cluster level, right?
also does read/write latency also come from cfstats result?

Posted 1 year ago #
thobbs
Administrator

Yes, nodetool displays stats for a single node, not for the cluster.

I'm not sure I understand your second questions, but if you are asking if OpsCenter gets its read/write latency numbers from the same place as nodetool cfstats, the answer is yes, but only for the column family specific latencies. The latencies on the dashboard are slightly different, as they are from the perspective of the coordinator node, rather than the node reading/writing data locally.

Posted 1 year ago #

RSS feed for this topic

12 Next »

Reply »

You must log in to post.