unstable
Cluster test: helpers/onlydots.tcl: detect EOF and exit.
Reply LOADING on replica while flushing the db (#13495)
Fix test "CLUSTER FORGET with invalid node ID" (#14421)
Cluster test: more reliable 01-faildet unit.
Cluster test: 02 unit more reliable waiting for slave sync.
Fix failing cluster tests (#9707)
Fix typos and replace 'codespell' with 'typos' (#15230)
Set repl-diskless-sync to yes by default, add repl-diskless-sync-max-replicas (#10092)
Fix replica count check in migration tests. (#10140)
Fixed some typos, add a spell check ci and others minor fix (#8890)
Cluster test: unit 09, Pub/Sub across the cluster.
Cluster: unit 10 modified to leave cluster in proper state.
Prevent replica failover during manual takeover test (#10499)
Rdb channel replication (#13732)
Change replica migration tests to use continous slots to improve speed (#10215)
Cluster: add test for the nofailover flag.
Handle primary/replica clients in IO threads (#14335)
Fix crash due to unblock client during slot migration (#13311)
Add reply_schema to command json files (internal for now) (#10273)
Test tcp deadlock fixes (#14946)
Fix brittle assert_match patterns for unexpected slowlog fields (#14948)
Temporarily hide the new SFLUSH command by marking it as experimental (#13600)
Revert cluster slot migration tests. (#8806)
tests: fix a typo (#9294)
Fix race condition in cluster test 22-replica-in-sync (#9721)
Track Pub/Sub subscription provenance across AUTH changes (#15468)
Fix flaky cluster pubsubshard test in 26-pubsubshard (#14962)
Add tests for error messages during slot migrations (#10449)