FAQ
What configuration is used for the master node?
The master node configuration is determined by the cluster operating mode.
- DEV: a virtual machine with 2 vCPU and 4 GB RAM will be used for the master node.
- PROD: a virtual machine with 6 vCPU and 6 GB RAM will be used for the master node.
Can I connect to the cluster using kubectl or a library?
To connect to the cluster, use access tokens in the VK Cloud management console or the Cloud ML Platform library. The token creation process is described in detail in the section Authentication and authorization.
How do I access the Spark History Server?
See the instructions.
How do I access a bucket?
By default, the Spark cluster has access to the VK Object Storage bucket, which is automatically created along with the cluster. To access other buckets, see the instructions.
Which monitoring and logging services can be connected to a Spark cluster?
Spark clusters are connected by default to the following services:
- Cloud Logging — log collection and analysis.
- Cloud Alerting — sending notifications about changes in key metrics of VK Cloud services.
- Cloud Monitoring — resource state tracking using a wide set of metrics.
For more information, see the article Cluster monitoring.
Was this article helpful?
Was this article helpful?