SOLR Cloud(8)Deployment and Backup Cloud

SOLR Cloud(8)Deployment and Backup Cloud

Since we already have EFS on the old cluster, we stop consumer and back up the data first
>http://alljobs1.us-east-1.elasticbeanstalk.com/solr/admin/collections?action=BACKUP&name=allJobsBackup18-01-17-16-39-00&collection=allJobs&location=/efs&async=123

Status check to make sure we already success on backing up
>http://alljobs1.us-east-1.elasticbeanstalk.com/solr/admin/collections?action=REQUESTSTATUS&requestid=123

You can also check the size in /efs directory
>du -h

Build the latest settings
>make ZK_ROOT=/solr/allJobsTest clean build

Push the latest Docker Image to Repository
>make push

Create a new cluster, and ask this cluster to join the same cloud as the previous cluster
>cd elasticbeanstalk/
>make ENVIRONMENT=allJobs2 ZK_ROOT=/solr/allJobs create

It the new machines did not join the cloud, we add replica
>curl "http://allJobsTest.us-east-1.elasticbeanstalk.com/solr/admin/collections?action=ADDREPLICA&collection=allJobs&shard=shard1"

After that the new cluster will copy the data from leader from the old cluster

Do Cname swap on AWS for the 2 clusters

After stable for 1 week, shutdown the old Cluster

References: