Release v1.0.132 is out 🚀
What's new:
- #1925 Implemented unjoin topics functionality for the p2p node.
- #1960 Epoch start optimization: make and register to only those topics for which you have information.
- #1961 Tweaked pool size capacity allocation.
Bug fixes:
- #1947 Fixed unjail nodes: jailed nodes remained jailed also after unjail, rating was reset but were not set to "new" list.
- #1952 Fixed trie snapshot concurrency: when
GetSnapshotThatContainsHash()is called, the pointer to thesnapshotDbis returned. But concurrently, the same db could have been closed and removed, thus resulting in a concurrency problem. - #1950 Fixed a stream error bug when the node deals with the elastic search indexer.
- #1957 Fixed rewards to metachain - make those invalid. Send all invalid rewards to community address.
- #1925 A peer that sends invalid message (that can not be unmarshaled) will be directly added to the black list.
- #1954 Added time.Context to the closing method on main.go as there is a known deadlock issue on P2P library when closing the process.
- #1956 Fixed snapshot order: The
trieStorageManagerdid not load snapshots in order. Each trie snapshot has an integer as directory name for the snapshot db. When the node loads the snapshots, they must be loaded in order (directory 9 before 10). - #1962 Fixed too many request when round info is indexed.
- #1965 Fixed state of meta stuck: fixed situation when shards will consider metachain stuck forever if the state would be set to stuck and because there was no way to reset this state back, as all the created blocks were empty. Shards should process txs cross me where they also would include new meta blocks, so the unstuck state of metachain could be restored.
Note: This release should start with a new DB