/cain

Cain is a backup and restore tool for Cassandra on Kubernetes

Primary LanguageGo

Cain

Cain is a backup and restore tool for Cassandra on Kubernetes. It is named after the DC Comics superhero Cassandra Cain.

Now an official part of the Helm incubator/cassandra chart!

Install

wget -qO- https://github.com/maorfr/cain/releases/download/0.1.0/cain.tar.gz | sudo tar xvz -C /usr/local/bin

Build from source

Cain uses glide as a dependency management tool, since some of the referenced packages are not available using dep.

glide up
go build -o cain cmd/cain.go

Commands

Backup Cassandra cluster to S3

Cain performs a backup in the following way:

  1. Backup the keyspace schema (using cqlsh) and copy it to S3.
  2. Get backup data using nodetool snapshot - it creates a snapshot of the keyspace in all Cassandra pods in the given namespace (according to selector).
  3. Copy the files in parallel to S3 using Skbn - it copies the files to the specified dst, under namespace/<cassandrClusterName>/keyspace/<keyspaceSchemaHash>/tag/.
  4. Clear all snapshots.

Usage

$ cain backup --help
backup cassandra cluster to S3

Usage:
  cain backup [flags]

Flags:
  -c, --container string   container name to act on (default "cassandra")
      --dst string         destination to backup to. Example: s3://bucket/cassandra
  -h, --help               help for backup
  -k, --keyspace string    keyspace to act on
  -n, --namespace string   namespace to find cassandra cluster
  -p, --parallel int       number of files to copy in parallel. set this flag to 0 for full parallelism (default 1)
  -l, --selector string    selector to filter on

Example

cain backup \
    -n default \
    -l release=cassandra \
    -k keyspace \
    --dst s3://db-backup/cassandra \
    -p 0

Restore Cassandra backup from S3

Cain performs a restore in the following way:

  1. Truncate all tables in keyspace.
  2. Copy files from the specified src (under keyspace/<keyspaceSchemaHash>/tag/) - restore is only possible for the same keyspace schema.
  3. Load new data using nodetool refresh.
  • Cain does not currently restore the schema (must be loaded restoring).

Usage

$ cain restore --help
restore cassandra cluster from S3

Usage:
  cain restore [flags]

Flags:
  -c, --container string   container name to act on (default "cassandra")
  -h, --help               help for restore
  -k, --keyspace string    keyspace to act on
  -n, --namespace string   namespace to find cassandra cluster
  -p, --parallel int       number of files to copy in parallel. set this flag to 0 for full parallelism (default 1)
  -l, --selector string    selector to filter on
      --src string         source to restore from. Example: s3://bucket/cassandra/namespace/cluster-name
  -t, --tag string         tag to restore

Example

cain restore \
    --src s3://db-backup/cassandra/default/ring01
    -n default \
    -k keyspace \
    -l release=cassandra \
    -t 20180903091624 \
    -p 0

Describe keyspace schema

Cain describes the keyspace schema using cqlsh. It can return the schema itself, or a checksum of the schema file (used by backup and restore).

Usage

$ cain schema --help
get schema of cassandra cluster

Usage:
  cain schema [flags]

Flags:
  -c, --container string   container name to act on (default "cassandra")
  -h, --help               help for schema
  -k, --keyspace string    keyspace to act on
  -n, --namespace string   namespace to find cassandra cluster
  -l, --selector string    selector to filter on
      --sum                print only checksum

Example

cain schema \
    -n default \
    -l release=cassandra \
    -k keyspace

cain schema \
    -n default \
    -l release=cassandra \
    -k keyspace \
    --sum

Support for additional storage services

Since Cain uses Skbn, adding support for additional storage services is simple. Read this post for more information.

Credentials

Kubernetes

Cain tries to get credentials in the following order:

  1. if KUBECONFIG environment variable is set - skbn will use the current context from that config file
  2. if ~/.kube/config exists - skbn will use the current context from that config file with an out-of-cluster client configuration
  3. if ~/.kube/config does not exist - skbn will assume it is working from inside a pod and will use an in-cluster client configuration

AWS

Skbn uses the default AWS credentials chain.

Examples

  1. Helm example
  2. Code example