v1.36.0

v1.36.0

Changes

Operator

• Documentation updates for fault domain and ADDITIONAL_ENV_FILE #1980
• Add additional documentation for the copied cluster-file #1978

v1.35.0

v1.35.0

Changes

Operator

• Remove an unnecessary error return value by compiling the regex #1977
• Improve documentation #1975
• Report correct current uptime in log line for safe bounce check #1974
• Correct version check for automatic tester process removal #1972
• Add safety checks for inclusion and improve the exclusion check #1971
• Bump google.golang.org/protobuf from 1.28.1 to 1.33.0 #1970
• Update the state of the helm charts #1968
• Initial changes for new maintenance mode integration #1967
• make GetProcessSettings not return volumeClaimTemplate for stateless … #1965
• Update documentation to mention the preferred nature of the generated pod anti-affinity rule and how to override it #1963
• Add the known limitation that the FDB cluster is only reachable from within Kubernetes #1961
• Allow more addresses to be present in the upgrade bounce case #1960
• Update pods check needs removal #1954

v1.34.0

v1.34.0

Changes

Operator

• Minor typo fix in mention of localities #1957
• Document issue with older fdb bindings and DNS support #1956
• Fix the e2e test case for the new FDB version that supports to remove old tester processes automatically #1955
• Fix nil pointer ha e2e #1952
• Disable flaky test in nightly runs #1947

v1.33.0

v1.33.0

Changes

Operator

• Allow to specify the storage engine when running e2e tests #1948
• Allow to specify a cluster name in test suite #1946
• Correct docs and improve log output #1945
• add podName selection functions #1944
• Add support for replacing failed process groups in a fault domain #1943
• Let the operator requeue another reconcile attempt if more failed process groups are detected #1942
• use common process selection function in buggify crash-loop, buggify no-schedule, and restart commands #1941
• Add condition selection option to common processGroup selection function #1940
• Only log if the command was not timing out #1939
• Don't skip the exclusion for process groups without an address when using localities #1938
• Create a unified process selection function and use it in removeProcessGroups #1937
• Improve verbosity of cordon and remove process commands #1936
• Allow the remove sub command to remove by process class #1935
• add ProcessGroupIdPrefix to testing generateClusterStruct and correct test cases #1934
• If the database is unavailable and caching is enabled, allow operator to proceed #1932
• Making the cluster argument optional when we are removing pods in the… #1930
• Remove the global admin mutex for the admin client #1929
• Reduce the update calls for the FoundationDB resource #1926
• Make sure we test the status conditions before running the configure command #1852

v1.32.0

v1.32.0

Changes

Operator

• Add estimates and pretty printing to exclusion status check #1927
• Allow the operator to restart the CC when tester processes are unreachable #1925
• Add additional safety checks for bouncing and excluding processes #1924
• Update the FDB versions used for e2e tests #1923
• Add cache client selector #1922
• Add additional tests for the maintenance mode #1921
• Make sure we exclude the FDB_NETWORK_OPTION_CLIENT_THREADS_PER_VERSION env variable and don't pass it down to fdbcli #1920
• Change base image to rocky9 #1894

v1.31.0

Release v1.31.0

v1.30.0

v1.30.0

Changes

Operator

• Enable 2 Storage Servers per default in test cases #1904
• Add the informer for getting a Pod by the node name #1903
• Document the logging fields that are commonly used in the operator logging statements #1898

v1.29.0

v1.29.0

Changes

Operator

• Fix logic when sidecar is unreachable for missing process detection #1896
• Update the way the operator deployment is created and verified #1895
• Add more logs when the fetched status has issues #1893
• Allow to set the namespace in the three data hall example #1892
• fixing error in AllAddressesExcluded #1891
• Add cache to plugin #1888
• Improve exclusion check to make sure that enough processes are up and running #1887
• Ignore HA locality-based exclusion test if version doesn't support locality based exclusions #1886
• Three data hall enable locking by default #1885
• Allow to define creationTimestamp mock #1882
• Add check for plugin version upgrades #1881
• Make sure the exclusion is tried again #1880
• Make sure the e2e test suite only runs locality based e2e tests if the provided versions support it #1879
• Refactor the check process method and remove some duplicate work #1870
• including processes using excluded server list #1857

v1.28.1

v1.28.1

Changes

Operator

• Fix database configuration bug if only proxies are specified #1883

v1.28.0

v1.28.0

Changes

Operator

• Make sure the operator checks all provided addresses for exclusions #1875
• Simplify test setup #1871
• Three data hall fault domain storage check bug fix #1869
• Correct the version check for the locality based exclusions #1868
• Make sure the operator can proceed with exclusions even if multiple pods are failing #1867
• Add some buffer for the exclusion logic to mitigate dead locks #1866
• Add support for redwood storage engine #1865
• Make use of string builder for getting the configuration string #1863
• Fix resources limits for non performance tests #1859
• If locality-based exclusions are enabled, only make use of the locality #1844