Reconstituting a Snowth Node

Warning:

This procedure is only used in circumstances where the node’s data is completely unrecoverable. Always contact Circonus Support (support@circonus.com) before attempting these procedures.

Building A Node Within The Same Cluster

This procedure is covered in detail in the IRONdb Operations manual. However, there are some important differences when performing this within the context of a Circonus Inside deployment:

  • No manual package installation is required. Hooper takes care of installing all of the necessary packages.
  • The service name is circonus-snowth.
  • Filesystem paths start with /snowth instead of /irondb, but descendant directory structure is the same.
    • When getting the name of the base dataset, use
      BASE_DATASET=`zfs list -H -o name /snowth`
      
  • Use the following command to start the reconstitute or to resume after an interruption:
    sudo /opt/circonus/bin/snowth-start -B
    

Building A Node In A New Cluster From The Old Cluster

This is a complex task. If you have questions or need assistance, please contact Circonus Support (support@circonus.com).

This procedure is covered in detail in the IRONdb Operations manual. However, there are some important differences when performing this within the context of a Circonus Inside deployment:

  • No manual package installation is required. Hooper takes care of installing all of the necessary packages.
  • The service name is circonus-snowth.
  • Filesystem paths start with /snowth instead of /irondb, but descendant directory structure is the same.
  • Use the following command to start the reconstitute or to resume after an interruption:
    sudo /opt/circonus/bin/snowth-start -B \
        -T <source_cluster_topo_hash> \
        -O <source_cluster_node_ip>:<port>