Branch 3.3

Branch 3.3

• need dashboard for Scylla Manager 2.1 #899
• Alternator logo #895
• prometheus.consul.yml.template add the manager agents target #888
• CQL Optimization dashboard gauges can be OFF for various time frames #881
• Shows speculative execution in official detail dashboard #880
• read timeout metrics are misleading #876
• Metrics to improve LWT dashboard #875
• We need Scylla OS 4.0 dashboard #871
• Hints derived-class metrics are not being displayed using irate() #865
• Default graphs hover tooltip to decreasing sort order #861
• Alternator option Alternator #858
• Add per group CPU information #857
• Add graphs for CDC #819
• Default Latency Alerts enhancement #803
• Grafana for Scylla 4.0 shows sextelion background reads per second #913
• views panel missing rate for counters #902

Branch 3.2

New in Branch 3.2

• Add scylla open source 3.3 dashboards #805
• Update the deafult version to 3.3,3.2 #849
• upgrade grafana to 6.6 #801
• upgrade prometheus to 2.15.1 #798
• Add Grafana annotation for Manager actions #815
• add an annotation for potentials stalls #649
• show all possible classes for I/O enhancement #839
• Allow overriding the default anonymous role for Grafana. #822
• Add ignore future metrics #809
• Add bypass cache to the CQL dashboard #806
• scylla-manager agent metrics' should be collected #795
• Add lightweight transaction metrics (aka CAS) to default dashboard #775
• add hints related graphs #545

Bug Fixes

• Where are the help strings? #841
• compaction shares graph is incorrect #817

Branch 3.1

New in Branch 3.1

• Scylla Open Source Version 3.2 dashboards
• Scylla Manager Version 2.0
• Grafana 6.5.1
• Prometheus 2.14
• Add the ad hoc filter to all dashboards #773
• Report-a-problem button #691
• Adding annotations for server restart #712
• CQL Optimization potential hazard with Consistency Level #706 #579
• Scylla manager consul integration #470

Bug Fixes

• Migrate single-stat panel to gauge #774
• Remove mountpoint from io queue #770
• 3.1 OS metrics - doesn't show correct disk in pie graphs #768
• MV section graphs are unrelated to MVs #775

Branch 3.0.4

Bug fix:

• Dashboards for scylla branch-3.2 don't display data #782

Branch 3.0.3

Adding Dashboards for Scylla 3.2

Branch 3.0.2

Bug Fixes

• #770 3.1 IO metrics Compactions I/O Queue
• #768 3.1 OS metrics - doesn't show correct disk in pie graphs
• #755 MV section graphs are unrelated to MVs

Branch 3.0.1

• Adding the alternator dashboard.
  To use the dashboard use the Scylla master version i.e. add -v master to the start-all.sh command.

More on Alternator can be found here

Branch 3.0

Notice

Please check the upgrade guide here

General Changes

• Dashboards rename (#700)
• Drop redundant labels (#687)
• Add a warning when the scylla_server.yml is missing (#685)
• Rename scylla_server and scylla_manager yml files to example (#657)
• Switch to Prometheus 2.10 (#639)
• switch to python 3 (#534)
• Switch to Grafana 6 (#526)
• Remove 2017.1, 1.5-2.3 dashboards from master (#634)
• Point users to the logs when a container fails to start (#647)
• Do not use port mapping when using the host network (#642)
• Switch to Prometheus 2.10 (#639)
• Allow adding a row to a dashboard (#621)

New alerts

• Add an alert on low disk space usage on root (#627)
• Add an alert when a node changes its status to leaving (#624)

Changes to the dashboards

• Task quote violation in the cpu dashboard should be by group (#651)
• Remove the node_exporter related metrics from the disk io dashboard (#645)
• Shorter version in the node table enhancement(#636)
• Update node_exporter metrics names to 0.17 (#633)
• Enhance node table with direct link per dashboard (#675)

Branch 2.4

New

• Adding Scylla 3.1 dashboards #594
• Adding the error Dashboard #576
• Add a graph for ad-hoc queries #610
• Allow changing aggregation modes #520
• State which metrics are coordinators and which replicas #521
• Allow external directory for grafana #597
• genconfig using nodetool improved #613
• Version per node table #151
• Add a warning on low disks pace on root partition #627
• Add alert when a node is leaving the cluster #623

Bug Fixes

• cql-optimization shows N/A for cross dc when there is no reads in a DC #608
• wrong metrics for compaction shares #604
• the group by shard is incorect #337
• scylla manager metric scylla_manager_cluster_cql_rtt_ms renamed #637
• Manager 1.4 dashboard is broken because of cql metrics rename #661
• cql optimization cross shard is off and should be remove #659

Scylla Branch 2.3.1

Bug fixes:

• Compactions I/O Queue delay by Shard should be filtered by mountpoint as well #588
• scylla manager panels not showing information when there are multiple keyspaces #602
• wrong metrics for compaction shares #604