Announcing ClusterVisor 1.23.08 with Major New Features

Posted on August 14, 2023

We are excited to announce the latest version of ClusterVisor (1.23.08) is now available.  This release not only adds many bug fixes reported by users, but a couple of major new features to stats / alerts, and stateless nodes. Here’s an overview of the new features:

Stats / Alerts:
ClusterVisor’s alerting mechanism has changed to be more comprehensive, and is now able to alert you on not only stats but other CV events.  The “red bell” icon in WebUI has been overhauled to make looking at and dealing with alerts much more manageable.  There is also new command line tools:

  • “cv-alerts” which will allow you to view and acknowledge any alerts from the command line.
  • “cv-health” that shows a summary of the health of CV itself, and can be helpful for identifying issues in the overall setup

In addition to alerting improvements, CV can now externally collect stats as well. This allows for monitoring when nodes go offline, the current state of the Slurm partitions, and also pulling stats from external devices via SNMP.  We currently support multiple types of PDUs, UPSs, and switches.  We expect to grow our support for new devices with customer feedback.  If there is a device you’d like CV to monitor that it currently doesn’t support, we want to hear from you at [email protected].

The last new stat feature is custom external stats. This is more of an advanced feature that allows you to inject your own stats into CV via a new command line tool “cv-write-stat”.  With this tool you can build your own cron job or other process to inject stats from any source you wish into CV.  Contact support if you would like to implement this on your own system.

Stateless Nodes:
Based on multiple customer requests we’ve completely re-engineered the stateless (diskless) image support in ClusterVisor.  The process is greatly simplified, and the same cloner image can now be deployed in both a stateful (installed on disks) or stateless (run completely from RAM) method.

This completely new rewrite doesn’t rely on NFS like previous versions, and is much easier to manage, and doesn’t require the same level of operating system changes as our previous versions.  If you are interested in stateless nodes, we’d love to hear your requirements and thoughts so we can continue to enhance and improve this feature set.

Future:
We have many other existing features planned for this year.  Keep an eye out for the newsletter, or subscribe to our ClusterVisor announce mailing list at https://lists.advancedclustering.com/mailman/listinfo/clustervisor-announce

We are always open to feedback, please contact [email protected] with your feature requests.

ClusterVisor cluster manager software

Download our HPC Pricing Guide
Get our Guide to Grant Writing

Request a Consultation from our team of HPC and AI Experts

Would you like to speak to one of our HPC or AI experts? We are here to help you. Submit your details, and we'll be in touch shortly.

  • This field is for validation purposes and should be left unchanged.