Upgrade Guide - Scylla 4.2 to 4.3 for Red Hat Enterprise Linux 7/8 or CentOS 7/8¶
This document is a step by step procedure for upgrading from Scylla 4.2 to Scylla 4.3, and rollback to 4.2 if required.
This guide covers upgrading Scylla from the following versions: 4.2.x or later to Scylla version 4.3.y, on the following platforms:
Red Hat Enterprise Linux 7/8 or CentOS 7/8
Execute the following commands one node at the time, moving to the next node only after the upgrade procedure completed successfully.
Upgrading your Scylla version is a rolling procedure that does not require a full cluster shutdown. For each of the nodes in the cluster, serially (i.e. one at a time), you will:
Check the cluster’s schema
Drain the node and backup the data
Backup the configuration file
Stop the Scylla service
Download and install new Scylla packages
Start the Scylla service
Validate that the upgrade was successful
Apply the following procedure serially on each node. Do not move to the next node before validating the node is up and running with the new version.
Once all the nodes have been upgraded to the new version 4.3, run a serial restart of all the nodes.
During the rolling upgrade, it is highly recommended:
Not to use new 4.3 features
Not to run administration functions, like repairs, refresh, rebuild or add or remove nodes. See here for suspending Scylla Manager (only available Scylla Enterprise) scheduled or running repairs.
Not to apply schema changes
Before upgrading, make sure to use Scylla Monitoring 3.4 or newer, for the Dashboards.
Make sure that all nodes have the schema synched prior to upgrade as any schema disagreement between the nodes causes the upgrade to fail.
Before any major procedure, like an upgrade, it is highly recommended to backup all the data to an external device. In Scylla, backup is done using the
nodetool snapshot command. For each node in the cluster, run the following command:
nodetool drain nodetool snapshot
Take note of the directory name that nodetool gives you, and copy all the directories having this name under
/var/lib/scylla to an external backup device.
When the upgrade is complete (for all nodes), remove the snapshot by running
nodetool clearsnapshot -t <snapshot>, or you risk running out of disk space.
Before upgrading, check what Scylla version you are currently running with
rpm -qa | grep scylla-server. You should use the same version as this version in case you want to rollback the upgrade. If you are not running a 4.2.x version, stop right here! This guide only covers 4.2.x to 4.3.y upgrades.
Update the Scylla rpm repo to 4.3
Install the new Scylla version
sudo yum update scylla\* -y
Alternator users upgrading from Scylla 4.0 to 4.1, need to set default isolation level
Check cluster status with
nodetool statusand make sure all nodes, including the one you just upgraded, are in UN status.
curl -X GET "http://localhost:10000/storage_service/scylla_release_version"to check the Scylla version. Validate that the version matches the one you upgraded to.
journalctl _COMM=scyllato check there are no new errors in the log.
Check again after two minutes, to validate no new issues are introduced.
Once you are sure the node upgrade was successful, move to the next node in the cluster.
Execute the following commands one node at the time, moving to the next node only after the rollback procedure completed successfully.
The following procedure describes a rollback from Scylla release 4.3.x to 4.2.y. Apply this procedure if an upgrade from 4.2 to 4.3 failed before completing on all nodes. Use this procedure only for the nodes that you upgraded to 4.3
Scylla rollback is a rolling procedure that does not require a full cluster shutdown. For each of the nodes rollback to 4.2, you will:
Drain the node and stop Scylla
Retrieve the old Scylla packages
Restore the configuration file
Reload the systemd configuration
Restart the Scylla service
Validate the rollback success
Apply the following procedure serially on each node. Do not move to the next node before validating the rollback was successful and that the node is up and running with the old version.
Remove the old repo file.
sudo rm -rf /etc/yum.repos.d/scylla.repo
Update the Scylla rpm repo to 4.2
sudo yum clean all sudo rm -rf /var/cache/yum sudo yum remove scylla\*tools-core sudo yum downgrade scylla\* -y sudo yum install scylla
sudo rm -rf /etc/scylla/scylla.yaml sudo cp -a /etc/scylla/scylla.yaml.backup-src| /etc/scylla/scylla.yaml
Restore all tables of system and system_schema from previous snapshot, 4.3 uses a different set of system tables. Reference doc: Restore from a Backup and Incremental Backup
cd /var/lib/scylla/data/keyspace_name/table_name-UUID/snapshots/<snapshot_name>/ sudo cp -r * /var/lib/scylla/data/keyspace_name/table_name-UUID/ sudo chown -R scylla:scylla /var/lib/scylla/data/keyspace_name/table_name-UUID/
Require to reload the unit file if the systemd unit file is changed.
sudo systemctl daemon-reload