Demo of manual failover of a cluster:
- cluster KV only initialized once, not on each restart of monit
- redis now logging to file instead of stdout
- confd can successfully reload redis & consul (it is now running as root)
make_manifest- specify name of deployment/service name with $NAME
Progress towards Redis cluster being able to failover and dynamically reconfigure itself.
- [consul] redis.conf and consul's redis.json created by confd (from consul kv) rather than directly from BOSH
- bug fix: Also kill redis server via pid
- consul performs health checks: available on port, disk 50% warning and 98% critical
- correctly use /var/vcap/store/redis for persistence
- ec2 templates add a persistent disk
- consul basic health check by writing to master node
- supports consul to advertise service
- upgrade templates/ to look like other boshreleases
- redis 2.8.3
- redis 2.6.13
- single VM only