The ScyllaDB Manager team is pleased to announce the release of ScyllaDB Manager 2.3, a production-ready version of ScyllaDB Manager for ScyllaDB Enterprise and ScyllaDB Open-Source customers. ScyllaDB Manager is a centralized cluster administration and recurrent tasks automation tool.
ScyllaDB Manager 2.3 brings a new suspend / resume commands, to be used for maintenance windows, as well as improvements in sctool, backup and monitoring.
ScyllaDB Enterprise customers are encouraged to upgrade to ScyllaDB Manager 2.3 in coordination with the ScyllaDB support team.
The new release includes upgrades of both ScyllaDB Manager Server, Manager CLI tool (sctool).
Useful Links
- Download ScyllaDB Manager and ScyllaDB Manager Agent for ScyllaDB Enterprise customers
- Download ScyllaDB Manager and ScyllaDB Manager Agent for open source users (up to 5 nodes)
- ScyllaDB Manager Docker instance and example
- ScyllaDB Manager 2.3 documentation
- Upgrade from ScyllaDB Manager 2.2 to ScyllaDB Manager 2.3
- Submit a ticket for questions or issues with Manager 2.3 (ScyllaDB Enterprise users)
New Features
Suspend and Resume Operations
ScyllaDB Manager can run multiple recurring tasks of different types (repair, backup) on multiple keyspaces and tables. In some cases, a cluster admin wants to temporarily suspend ALL of the tasks in favor of a planned maintenance window. Until now, the only way to do so was to stop the tasks one by one.
In ScyllaDB Manager 2.3, we are adding two new commands suspend
and resume
to do that. The command can be executed from the CLI, using sctool, or using the REST API.
You can use these APIs to integrate with an external scheduler, to set a maintenance window in advance.
Consult the online docs for reference.
Backup Improvements
ScyllaDB Manager includes recurrent backup tasks to S3 and GCS. (Read the documentation.)
In ScyllaDB Manager 2.3 the following improvement have been implemented:
Backup Manifest
Backup manifest is a file included in each backup with information on the backup content.
The following updates to the manifest are included in Manager 2.3:
- Manifest will include the cluster name and IP address of a node.
- Replace keyspace token ranges with tokens as in system.local.tokens
The manifest is backward compatible, so Manager 2.3 can read older versions manifest, but not vice versa.
Stability and Debugging
- ScyllaDB Manager has a protection logic from resuming old, out of date backup tasks. This release fixes a bug in this logic, evaluating the backup age max against time from snapshot tag.
- Backup gracefully fails when indexing finds no snapshot files (removed by another process).
- If ScyllaDB Manager Agent or Server failed to start due to wrong configuration there was no indication of that in the logs. Now journalctl will show the error message.
- If you had multiple clusters added and one of them was down sctool status and other commands would fail, now they work for the running clusters.
- Grace period is removed from the scheduler if you schedule something to run now it would run immediately it hits the server.
Improvements to backup logging.
Monitoring
New metrics in Manager 2.3
- A metric for the last success of a task. The metric
scylla_manager_task_last_success
records start time of the last successful run as a Unix timestamp. - If service was killed after restart the
scylla_manager_task_run_total
metric with statusABORTED
is set to 1 for every aborted task.
You can use ScyllaDB Monitoring Stack 3.6.3 with -M 2.3
option to get Manager 2.3 dashboards.
25 Mar 2021