Bug Fixes in Release 4.4.3

• Fix: Some manager repair/backup timestamps are missing #2034
• Fix: Alerts severities are inverted #2029

This is a security patch release.

Following Grafana CVE-2023-3128 announcement.

The release contains an upgrade to Grafana version 9.5.5

Release 4.4.1

This is a security patch release.

Following Grafana CVE-2023-2183 and CVE-2023-2801 announcement.

The release contains an upgrade to Grafana version 9.5.3

New in 4.4.0

• Scylla Manager: Add Backup restore remaining bytes panel #1962
• Update the Advisor section #1953
• Add a scheduling_group_name filter to the overview dashboard #1940
• Remove internal scheduling_group_name from p99 in the overview and allow choosing the relevant SG #1939
• Add panels for CQL request and response sizes #1928
• Disk size panel update #1958
• Show scylla_io_queue_starvation_time_sec metrics #1915
• Add a command line option to drop cas/cdc metrics #1948
• Add missing descriptions to panels #1966
• Update the alerts table #1930

Bug Fixes

• Empty alert manager rules when using consul #1951
• The links in the advisor balance section are broken #1949
• Fix legend of IO-Queue panels #1916

Operational changes

• Prepare for summaries #1972
• Reduce the number of collected histograms: do not collect "internal" scheduling groups' latency histograms #1971
• Datadog - do not report per-shard metrics #1944
• Change alerts level to text #1927

• Support Scylla Manager 3.1

Security Release

Following Grafana Security Release cve-2023-1410 This patch release update Grafana version to 9.3.11

Bug Fixes in 4.3.1

• Gaps in recording rules graphs in clusters with many cores #1935

New in Release 4.3.0

• Ready for 5.2 #1881
• Rename the datadog labels #1902
• Remove the alertmanger plugin #1898
• Allow selecting specific CPUs on OS dashboard #1878
• Use p50 quantile instead of average for latencies #1889
• More robust alert for node reports being up, while going down. #1877
• Allow admin password and an anonymous viewer #1913
• Add a CPU starvation time graph in the per-scheduling-group section #1922
• Tooltip in desc order under the avanced dashboard #1916

Bug Fixes in Release 4.3.0

• Scylla CQL instance all filter should list all nodes #1895
• prometheus.consul.yml: take the cluster name from the consul #1891
• scylla-detailed rename the active and queued read correctly #1876
• Different value of metric of calculate by recording rules of prometheus and on-fly calculate #1815

Bug Fixes

• Rename the active and queued reads in the detailed dashboard #1876
• The all_scheduling_group filter should use both read and write and be per cluster #1872
• Fallback rules in per scheduling groups latency should use the scheduling-group filter #1871
• Annotations should be limited to the current cluster #1866

New in Scylla Monitoring 4.2.0

• scylla-detailed: keep the latency pannels open but don't show empty ones #1854
• Mark base level #1853
• versions.sh move prometheus to version 2.41 #1846
• Set Grafana version to 9.3.4 #1864
• scylla-advanced.template.json: Split the disk and io-queue panels #1827

Operational Change

• kill-all.sh gracefully kill prometheus with sigterm #1832
• Added Recording rules for Datadog integration #1855
• Optionally use VictoriaMetrics instead of Prometheus - Experimental #1542
• load-grafana.sh generate dashboards from the new naming convention #1849

Bug Fixes

• types.json: set the all option as .* #1840
• scylla manager 3.0 metrics update #1847