Issue to start coordinator - arangodb

I have install 3 arangodb servers. But i have always the same listening port 8529 no 8530 for coordinator so i cannot create a cluster.
tcp 0 0 0.0.0.0:8529 0.0.0.0:* LISTEN 13142/arangod
So when i try to create a cluster via the web interface, i have the following error
ERROR bootstrapping DB servers failed: Could not connect to 'tcp://10.0.0.18:8530' 'connect() failed with #111 - Connection refused'
How can i start and/or configure the corrdinator to have a listen on my servers?
Regards

Dispatcher based clusters
Please note that dispatcher based setups as you asked are intended for evaluation purposes only.
To start a cluster from the dispatcher webfrontend you need to configure all nodes to start the arangod daemon in dispatcher mode:
[cluster]
disable-dispatcher-kickstarter = no
disable-dispatcher-frontend = no
To start a cluster on a single machine you only need to install ArangoDB and reconfigure it once; it will then use the same installation to start the dispatcher and dbserver nodes.
One should know that the initial cluster startup may take a while.
Another side note is that authentication is not supported in this scenario, so you may need to turn it off.
You should now find the log output of the dbserver and coordinator instances under /var/log/arangodb/cluster/ so you can get the actual informations of what went wrong.
Script based cloud install clusters
A better way to get a cluster running in the cloud may be to use one of the scripts we prepared for Digital Ocean, Google Compute Engine, AWS or Azure.
ArangoDB Clusters based on Mesosphere DCOS
The currently recommended way of running an ArangoDB cluster is to use Mesosphere DCOS, as Max describes in these slides using some example configurations.
ArangoDB is an official Mesosphere partner and we offer an official DCOS subcommand to manage an ArangoDB Cluster on Mesosphere DCOS.
Mesosphere adds additional services on top of Mesos and eases management of the Mesos cluster via the dcos-cli.
If you want to use a raw Apache Mesos Cluster, you can use the Mesos framework directly to schedule the neccessary tasks to create ArangoDB cluster.
Meanwhile there is a better article about Running ArangoDB on DC/OS.

Related

How to run a Spark Standalone master on Kubernetes that will use the Kubernetes Cluser Manager to start workers

I have an application that currently uses Standalone Mode locally to use spark functionality via the SparkContext. We are not using spark-submit to upload our jobs, we are running our application in a container on kubernetes so we would like to take advantage of the dynamic scheduling that kubernetes provides to run the jobs.
We started out looking for a helm chart to create stand alone cluster running on kubernetes similar to how you would have run a standalone cluster on machines ( vms or actual machines ) a few years ago and came across the following
https://github.com/helm/charts/tree/master/stable/spark
Issues:
very old instances of spark
not using the containers provided by spark
this setup wastes a bunch of resources if you need to have large worker nodes reserved and running all the time regardless of your need
Next we started looking at the spark-operator approach here https://github.com/GoogleCloudPlatform/spark-on-k8s-operator
Issues:
Doesn't support the way we interact with spark, takes the approach that all the apps are standalone apps that are pushed to the cluster to run
No longstanding master that allows us to take advantage of cached resources in the cluster
Along this journey we discovered that spark now supports a kubernetes cluster manager ( similar to the way it does with yarn, mesos ) so we are looking that this might be the best approach, but this still does not provide a standalone master that would allow for the in memory caching. I have looked to see if there was a way that I could get the org.apache.spark.deploy.master.Master to start and use the
org.apache.spark.scheduler.cluster.k8s.KubernetesClusterManager
So I guess what I'm trying to ask is does anyone have any experience in trying to run a Standalone Master, that would use the kubernetes backend such as "KubernetesClusterManager" in order to have the worker nodes dynamically created as pods and running executors while having a permanent Standalone Master that would allow a SparkContext to connect to it remotely in client mode.

Thingsboard cluster setup

Building a Thingsboard cluster
I need help setting up a Thingsboard cluster, the documentation online is very limited.
The cluster will contain 2 Zookeeper nodes and 4 Thingsboard nodes with Cassandra DB.
Should Zookeeper be installed separately?
A step-by-step guide would be much appreciated!
I cannot provide you detailed step-by-step instructions to setup a ThingsBoard cluster. I can point you into the right direction by sharing the different documents you need to do so.
Bottom line, the following tasks must be completed:
Install and configure a ZooKeeper ensemble.
Check the ZooKeeper documentation for further installation details. Keep in mind that you need at least three different ZK-nodes in a clustered environment and that you always need an odd number of ZK nodes (3,5,7,...). It is a very very very bad idea to build a cluster consisting out of two ZK-nodes, check split brain condition that might appear under these circumstances! Basically you setup the number of individual nodes you wish to use and change the configuration file to enable the different nodes as an ensemble. This is documented quite well in the ZK-docs.
Install and configure a Cassandra cluster.
Again you will setup the number of individual nodes you need for your Cassandra cluster and modify the individual configuration files to convert them into a Cassandra cluster. Check Cassandra documentation for details. Be sure to check proper configuration using the nodetool status command as described at the end of the document. All your nodes should be up and running.
Install and configure a ThingsBoard cluster.
Use the instructions provided with ThingsBoard single node setup.
Install Java
Skip External database installation
ThingsBoard service installation
Configure ThingsBoard to use the external database - Cassandra
Go to Cluster setup and apply the configuration steps depicted (ZK, Cassandra and RPC). Keep in mind to point to ALL members of your ZK, Cassandra cluster. You can also use IP-addresses instead of host names.
Return to single node setup and run the installation script at ONE NODE only!
Start ThingsBoard service
If everything went well, you should be able to access your ThingsBoard nodes directly using the URL http://[NODE_IP]:8080. You can verify proper cluster operation by creating a tenant on one node and check its presence on another node.
I don't know if using an even number of ThingsBoard nodes is a good idea. The documentation does not mention anything about this.
One final remark, you could/should consider putting a proxy in front of your ThingsBoard cluster to provide load balancing to your web clients and improve user experience. This way you shouldn't share the individual host addresses with your users and you will prevent node overloading due to the fact that everybody is using the same web-address to access your dashboard(s). You could also proxy your MQTT broker to provide load balancing as well.
Good luck in setting up your cluster!
Zookeeper needs at least 3 nodes to run in a cluster mode. Each node voting and the valid replica count to gain the QUORUM is 3.

Apache Cassandra monitoring

What is the best way to monitor if cassandra nodes are up? Due to security reasons JMX and nodetool is out of question. I have cluster metrics monitoring via Rest Api, but I understand that even if a node goes Rest Api will only report on a whole cluster.
Well, I have integrated a system where I can monitor all the metrics regarding to my cluster of all nodes. This seems like complicated but pretty simple to integrate. You will need the following components to build up a monitoring system for cassandra:
jolokia jar
telegraf
influxdb
grafana
I'm writing a short procedure, how it works.
Step 1: copy jolokia jvm jar to install_dir/apache-cassandra-version/lib/ , jolokia jvm agent can be downloaded from anywhere in google.
Step 2: add the following line to install_dir/apache-cassandra-version/conf/cassandra-env.sh
JVM_OPTS="$JVM_OPTS -javaagent:<here_goes_the_path_of_your_jolokia_jar>"
Step 3: install telegraf on each node and configure the metrics you want to monitor. and start telegraf service.
Step 4: install grafana and configure your ip, port, protocol. grafana will give you a dashboard to look after your nodes and start grafana service. Your metrics will be able get visibility here.
Step 5: install influxdb on another server from where you want to store your metrics data which will come through telegraf agent.
Step 6: browse the ip you have mentioned, where you have launched your grafana through browser and add data source ip (influxdb ip), then customize your dashboard.
image source: https://blog.pythian.com/monitoring-cassandra-grafana-influx-db/
This is not for monitoring but only for node state.
Cassandra CQL driver provides info if a particular node is UP or DOWN with Host.StateListener Interface. This info is used by driver to mark a node UP or Down. Thus it could be used if node is down or up if JMX is not accessible.
Java Doc API : https://docs.datastax.com/en/drivers/java/3.3/
I came up with a script which listens for DN nodes in the cluster and reports it to our monitoring setup which is integrated with pagerduty.
The script runs on one of our nodes and executes nodetool status every minute and reports for all down nodes.
Here is the script https://gist.github.com/johri21/87d4d549d05c3e2162af7929058a00d1
[1]:

How to setup Spark with a multi node Cassandra cluster?

First of all, I am not using the DSE Cassandra. I am building this on my own and using Microsoft Azure to host the servers.
I have a 2-node Cassandra cluster, I've managed to set up Spark on a single node but I couldn't find any online resources about setting it up on a multi-node cluster.
This is not a duplicate of how to setup spark Cassandra multi node cluster?
To set it up on a single node, I've followed this tutorial "Setup Spark with Cassandra Connector".
You have two high level tasks here:
setup Spark (single node or cluster);
setup Cassandra (single node or cluster);
This tasks are different and not related (if we are not talking about data locality).
How to setup Spark in Cluster you can find here Architecture overview.
Generally there are two types (standalone, where you setup Spark on hosts directly, or using tasks schedulers (Yarn, Mesos)), you should draw upon your requirements.
As you built all by yourself, I suppose you will use Standalone installation. The difference between one node is network communication. By default Spark runs on localhost, more commonly it uses FQDNS name, so you should configure it in /etc/hosts and hostname -f or try IPs.
Take a look at this page, which contains all necessary ports for nodes communication. All ports should be open and available between nodes.
Be attentive that by default Spark uses TorrentBroadcastFactory with random ports.
For Cassandra see this docs: 1, 2, tutorials 3, etc.
You will need 4 likely. You also could use Cassandra inside Mesos using docker containers.
p.s. If data locality it is your case you should come up with something yours, because nor Mesos, nor Yarn don't handle running spark jobs for partitioned data closer to Cassandra partitions.

How to run PySpark (possibly in client mode) on Mesosphere cluster?

I am trying to run a PySpark job on a Mesosphere cluster but I cannot seem to get it to run. I understand that Mesos does not support cluster deploy mode for PySpark applications and that it needs to be run in client mode. I believe this is where the problem lies.
When I try submitting a PySpark job I am getting the output below.
... socket.hpp:107] Shutdown failed on fd=48: Transport endpoint is not connected [107]
I believe that a spark job running in client mode needs to connect to the nodes directly and this is being blocked?
What configuration would I need to change to be able to run a PySpark job in client mode?
When running PySpark in client mode (meaning the driver is running where you invoke Python) the driver becomes the Mesos Framework. When this happens, the host the framework is running on needs to be able to connect to all nodes in the cluster, and they need to be able to connect back, meaning no NAT.
If this is indeed the cause of your problems, there are two environment variables that might be useful. If you can get a VPN in place, you can set LIBPROCESS_IP and SPARK_LOCAL_IP both to the IP of the host machine that cluster nodes can use to connect back to the driver.

Resources