8e0c89b76b
We need to trigger reconfiguring after purging some of the servers. Otherwise, if the customer has 2 standby nodes and they switch back to single server (only primary) deployment, primary server keeps waiting for write acknowledgement from standbys since we don't configure it.
64 lines
2.2 KiB
Ruby
64 lines
2.2 KiB
Ruby
# frozen_string_literal: true
|
|
|
|
require_relative "../../lib/util"
|
|
|
|
class Prog::Postgres::ConvergePostgresResource < Prog::Base
|
|
subject_is :postgres_resource
|
|
|
|
label def start
|
|
register_deadline("recycle_representative_server", 2 * 60 * 60)
|
|
hop_provision_servers
|
|
end
|
|
|
|
label def provision_servers
|
|
hop_wait_servers_to_be_ready if postgres_resource.has_enough_fresh_servers?
|
|
|
|
if postgres_resource.servers.all? { it.vm.vm_host } || postgres_resource.location.provider == "aws"
|
|
exclude_host_ids = []
|
|
if !(Config.development? || Config.is_e2e) && postgres_resource.location.provider == HostProvider::HETZNER_PROVIDER_NAME
|
|
used_data_centers = postgres_resource.servers.map { it.vm.vm_host.data_center }.uniq
|
|
exclude_host_ids = VmHost.where(data_center: used_data_centers).map(&:id)
|
|
end
|
|
Prog::Postgres::PostgresServerNexus.assemble(resource_id: postgres_resource.id, timeline_id: postgres_resource.timeline.id, timeline_access: "fetch", exclude_host_ids: exclude_host_ids)
|
|
end
|
|
|
|
nap 5
|
|
end
|
|
|
|
label def wait_servers_to_be_ready
|
|
hop_provision_servers unless postgres_resource.has_enough_fresh_servers?
|
|
|
|
hop_recycle_representative_server if postgres_resource.has_enough_ready_servers?
|
|
|
|
nap 60
|
|
end
|
|
|
|
label def recycle_representative_server
|
|
if (rs = postgres_resource.representative_server)
|
|
hop_prune_servers unless rs.needs_recycling?
|
|
|
|
nap 10 * 60 unless postgres_resource.in_maintenance_window?
|
|
|
|
register_deadline(nil, 10 * 60)
|
|
rs.trigger_failover
|
|
end
|
|
|
|
nap 60
|
|
end
|
|
|
|
label def prune_servers
|
|
# Below we only keep servers that does not need recycling. If there are
|
|
# more such servers than required, we prefer ready and recent servers (in that order)
|
|
servers_to_keep = postgres_resource.servers
|
|
.reject { it.representative_at || it.needs_recycling? }
|
|
.sort_by { [(it.strand.label == "wait") ? 0 : 1, Time.now - it.created_at] }
|
|
.take(postgres_resource.target_standby_count) + [postgres_resource.representative_server]
|
|
(postgres_resource.servers - servers_to_keep).each.each(&:incr_destroy)
|
|
|
|
servers_to_keep.each(&:incr_configure)
|
|
postgres_resource.incr_update_billing_records
|
|
|
|
pop "postgres resource is converged"
|
|
end
|
|
end
|