View Source Changelog

0.10.0

• Added optional TTL to Horde.DynamicSupervisor's :proxy_operation messages. The Time-to-Live defaults to :infinity for full backwards compatibility. This TTL helps prevent potential issues where messages could loop forever between a set of nodes which disagree on which node should execute the task.
• [BREAKING] Horde.DynamicSupervisor's new :proxy_message_ttl option configures the maximum TTL for proxy messages. It takes an integer denoting the maximum number of hops a message can travel, or the atom :infinity (default). This can be a breaking change: when upgrading do not set this option to an integer. You can explicity set it to :infinity or leave it default. If this is set to an integer, upgraded nodes won't be able to proxy to non-upgrade nodes.

0.9.1

• Fix race condition in registry when node disconnects
• Pass extra_arguments flag to the ProcessSupervisor
• Updating libring dependency to ~> 1.7. Needed for upgrade to OTP 27. See this PR to libring for details.

0.9.0

• Bugfixes for scenarios causing Horde to crash. See #266 and #263.
• [BREAKING] The first parameter of Horde.DistributionStrategy.choose_node/2 has changed from the identifier to the full child spec. See #239.
• Use :erpc instead of :rpc. See #265.
• Stop eagerly registering processes in Horde.Registry. This solves #250.
• Add Horde.UniformRandomDistribution process distribution strategy. See #252.

0.8.7

• Tweak dependency on :telemetry

0.8.6

• Fix an issue with members: :auto

0.8.5

• Add support for telemetry version 1.0.0

0.8.4

• Upgrade to delta_crdt 0.6.0

0.8.3

• Fix a deadlock that occurred if a process was being restarted while Horde.DynamicSupervisor.start_child/2 was being called. #218
• Respect max_restarts and max_seconds when given as options to Horde.DynamicSupervisor.child_spec/1. #216

0.8.2

• Bump version of telemetry_poller dependency. #212

0.8.1

• Horde.Registry.delete_meta/2 has been added to reflect its addition in Elixir.Registry in upcoming release 1.11.0 #208

0.8.0

• Horde.DynamicSupervisor behaviour in a netsplit has changed. Previously, when a netsplit heals, Horde.DynamicSupervisor would try to clean up any duplicate processes. It no longer does this, leaving that responsibility to Horde.Registry. #196
• Horde.DynamicSupervisor and Horde.Registry now support the option members: :auto to automatically detect other identically-named supervisors or registries. #184
• Horde.DynamicSupervisor now supports the option process_redistribution: :active to rebalance processes actively (aka, when a node joins or leaves the cluster). The default is :passive, which only redistributes processes when a node dies or loses quorum. #164.

0.7.1

• Use MFA for on_diff instead of anonymous function, avoids passing around functions (which can be error-prone). #167.

0.7.0

• Horde.Supervisor has been renamed to Horde.DynamicSupervisor.
• Horde.Registry and Horde.Supervisor now follow the api of Elixir.DynamicSupervisor more closely (specifically init/1 callback and module-based Supervisor / Registry). #152.
• Horde.Registry.lookup/2 returns [] instead of :undefined when no match. #145
• child_spec/1 can be overridden in Horde.Registry and Horde.Supervisor #135 #143
• Implement :listeners option for Horde.Registry. #142
• Fix via tuple usage with meta. #139

0.6.1

• Module-based Horde.Supervisor can override child_spec/1. #135
• Added guides for handling clustering, process state handoff (during deploys), and special considerations for eventual consistency to the documentation.
• Horde.Supervisor now uses libring to distribute processes over nodes. #130
• Horde.Supervisor publishes metrics with :telemetry ([:horde, :supervisor, :supervised_process_count]). #132
• Horde.Supervisor and Horde.Registry now support option delta_crdt_options, which you can use to tune your cluster. Also updated to the most recent DeltaCRDT. #100

0.6.0

• Horde.Supervisor now behaves more like DynamicSupervisor. #122
• Horde.Registry sends an exit signal to the process that "loses" when a conflict is resolved. #118
• Horde.Registry.register/3 returns {:error, {:already_registered, pid}} when applicable. This improves compatibility with Elixir.Registry. #115
• Adds Horde.Registry.select/2, which works the same as Elixir.Registry.select/2, which will land in Elixir 1.9. #110
• Fixes a bug causing Horde.Supervisor to crash if a child process was restarting when Horde.Supervisor.delete_child/2 was called. #114