4 Commits

Author SHA1 Message Date
anonpenguin23
dd4cb832dc
Add memory leak prevention in pubsub handler and improve database recovery logic
- Implemented a cleanup goroutine in the pubsub handler to remove old entries from the recent connections map, preventing memory leaks.
- Enhanced the database recovery process in the cluster manager by adding checks for quorum and ensuring proper handling of existing data during instance startup.
- Updated instance management to clear Raft state for leaders with existing cluster state and added background SQL readiness checks for followers.
- Improved logging for better visibility during recovery and instance startup processes.
2025-10-16 10:10:33 +03:00
anonpenguin23
94248e40b0
Updated default advertise address to 0.0.0.0 in main.go, removed unnecessary BEGIN/COMMIT statements from migration SQL files, and improved database initialization handling in cluster manager. Increased wait times for node readiness and SQL availability checks to 60 seconds. 2025-10-14 13:59:36 +03:00
anonpenguin23
0bd9749aaf
Made clustering with system db work and replicate but migrations do not run 2025-10-14 12:59:24 +03:00
anonpenguin23
f2d5a0790e
started working on clustering 2025-10-13 07:41:46 +03:00