status-go/services/ext/mailservers/connmanager_test.go

401 lines
11 KiB
Go
Raw Permalink Normal View History

Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
package mailservers
import (
"fmt"
"sync"
"testing"
"time"
2020-01-02 09:10:19 +00:00
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
"github.com/ethereum/go-ethereum/event"
"github.com/ethereum/go-ethereum/p2p"
"github.com/ethereum/go-ethereum/p2p/enode"
2020-01-02 09:10:19 +00:00
"github.com/status-im/status-go/eth-node/types"
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
"github.com/status-im/status-go/t/utils"
)
type fakePeerEvents struct {
mu sync.Mutex
nodes map[types.EnodeID]struct{}
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
input chan *p2p.PeerEvent
}
func (f *fakePeerEvents) Nodes() []types.EnodeID {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
f.mu.Lock()
rst := make([]types.EnodeID, 0, len(f.nodes))
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
for n := range f.nodes {
rst = append(rst, n)
}
f.mu.Unlock()
return rst
}
func (f *fakePeerEvents) AddPeer(node *enode.Node) {
f.mu.Lock()
f.nodes[types.EnodeID(node.ID())] = struct{}{}
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
f.mu.Unlock()
if f.input == nil {
return
}
f.input <- &p2p.PeerEvent{
Peer: node.ID(),
Type: p2p.PeerEventTypeAdd,
}
}
func (f *fakePeerEvents) RemovePeer(node *enode.Node) {
f.mu.Lock()
delete(f.nodes, types.EnodeID(node.ID()))
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
f.mu.Unlock()
if f.input == nil {
return
}
f.input <- &p2p.PeerEvent{
Peer: node.ID(),
Type: p2p.PeerEventTypeDrop,
}
}
func newFakePeerAdderRemover() *fakePeerEvents {
return &fakePeerEvents{nodes: map[types.EnodeID]struct{}{}}
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
}
func (f *fakePeerEvents) SubscribeEvents(output chan *p2p.PeerEvent) event.Subscription {
return event.NewSubscription(func(quit <-chan struct{}) error {
for {
select {
case <-quit:
return nil
case ev := <-f.input:
// will block the same way as in any feed
output <- ev
}
}
})
}
func newFakeServer() *fakePeerEvents {
srv := newFakePeerAdderRemover()
srv.input = make(chan *p2p.PeerEvent, 20)
return srv
}
type fakeEnvelopeEvents struct {
input chan types.EnvelopeEvent
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
}
func (f *fakeEnvelopeEvents) SubscribeEnvelopeEvents(output chan<- types.EnvelopeEvent) types.Subscription {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
return event.NewSubscription(func(quit <-chan struct{}) error {
for {
select {
case <-quit:
return nil
case ev := <-f.input:
// will block the same way as in any feed
output <- ev
}
}
})
}
2019-01-21 14:00:10 +00:00
func newFakeEnvelopesEvents() *fakeEnvelopeEvents {
return &fakeEnvelopeEvents{
input: make(chan types.EnvelopeEvent),
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
}
}
func fillWithRandomNodes(t *testing.T, nodes []*enode.Node) {
var err error
for i := range nodes {
nodes[i], err = RandomNode()
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
require.NoError(t, err)
}
}
func getMapWithRandomNodes(t *testing.T, n int) map[types.EnodeID]*enode.Node {
nodes := make([]*enode.Node, n)
fillWithRandomNodes(t, nodes)
return nodesToMap(nodes)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
}
func mergeOldIntoNew(old, new map[types.EnodeID]*enode.Node) {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
for n := range old {
new[n] = old[n]
}
}
func TestReplaceNodes(t *testing.T) {
type testCase struct {
description string
old map[types.EnodeID]*enode.Node
new map[types.EnodeID]*enode.Node
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
target int
}
for _, tc := range []testCase{
{
"InitialReplace",
getMapWithRandomNodes(t, 0),
getMapWithRandomNodes(t, 3),
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
2,
},
{
"FullReplace",
getMapWithRandomNodes(t, 3),
getMapWithRandomNodes(t, 3),
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
2,
},
} {
t.Run(tc.description, func(t *testing.T) {
peers := newFakePeerAdderRemover()
state := newInternalState(peers, tc.target, 0)
state.replaceNodes(tc.old)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
require.Len(t, peers.nodes, len(tc.old))
for n := range peers.nodes {
require.Contains(t, tc.old, n)
}
state.replaceNodes(tc.new)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
require.Len(t, peers.nodes, len(tc.new))
for n := range peers.nodes {
require.Contains(t, tc.new, n)
}
})
}
}
func TestPartialReplaceNodesBelowTarget(t *testing.T) {
peers := newFakePeerAdderRemover()
old := getMapWithRandomNodes(t, 1)
new := getMapWithRandomNodes(t, 2)
state := newInternalState(peers, 2, 0)
state.replaceNodes(old)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
mergeOldIntoNew(old, new)
state.replaceNodes(new)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
require.Len(t, peers.nodes, len(new))
}
func TestPartialReplaceNodesAboveTarget(t *testing.T) {
peers := newFakePeerAdderRemover()
initial, err := RandomNode()
require.NoError(t, err)
old := nodesToMap([]*enode.Node{initial})
new := getMapWithRandomNodes(t, 2)
state := newInternalState(peers, 1, 0)
state.replaceNodes(old)
state.nodeAdded(types.EnodeID(initial.ID()))
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
mergeOldIntoNew(old, new)
state.replaceNodes(new)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
require.Len(t, peers.nodes, 1)
}
func TestConnectionManagerAddDrop(t *testing.T) {
server := newFakeServer()
whisper := newFakeEnvelopesEvents()
target := 1
2019-01-21 14:00:10 +00:00
connmanager := NewConnectionManager(server, whisper, target, 1, 0)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
connmanager.Start()
defer connmanager.Stop()
nodes := []*enode.Node{}
for _, n := range getMapWithRandomNodes(t, 3) {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
nodes = append(nodes, n)
}
// Send 3 random nodes to connection manager.
connmanager.Notify(nodes)
var initial enode.ID
// Wait till connection manager establishes connection with 1 peer.
require.NoError(t, utils.Eventually(func() error {
nodes := server.Nodes()
if len(nodes) != target {
return fmt.Errorf("unexpected number of connected servers: %d", len(nodes))
}
initial = enode.ID(nodes[0])
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
return nil
}, time.Second, 100*time.Millisecond))
// Send an event that peer was dropped.
select {
case server.input <- &p2p.PeerEvent{Peer: initial, Type: p2p.PeerEventTypeDrop}:
case <-time.After(time.Second):
require.FailNow(t, "can't send a drop event")
}
// Connection manager should establish connection with any other peer from initial list.
require.NoError(t, utils.Eventually(func() error {
nodes := server.Nodes()
if len(nodes) != target {
return fmt.Errorf("unexpected number of connected servers: %d", len(nodes))
}
if enode.ID(nodes[0]) == initial {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
return fmt.Errorf("connected node wasn't changed from %s", initial)
}
return nil
}, time.Second, 100*time.Millisecond))
}
func TestConnectionManagerReplace(t *testing.T) {
server := newFakeServer()
whisper := newFakeEnvelopesEvents()
target := 1
2019-01-21 14:00:10 +00:00
connmanager := NewConnectionManager(server, whisper, target, 1, 0)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
connmanager.Start()
defer connmanager.Stop()
nodes := []*enode.Node{}
for _, n := range getMapWithRandomNodes(t, 3) {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
nodes = append(nodes, n)
}
// Send a single node to connection manager.
connmanager.Notify(nodes[:1])
// Wait until this node will get connected.
require.NoError(t, utils.Eventually(func() error {
connected := server.Nodes()
if len(connected) != target {
return fmt.Errorf("unexpected number of connected servers: %d", len(connected))
}
if types.EnodeID(nodes[0].ID()) != connected[0] {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
return fmt.Errorf("connected with a wrong peer. expected %s, got %s", nodes[0].ID(), connected[0])
}
return nil
}, time.Second, 100*time.Millisecond))
// Replace previously sent node with 2 different nodes.
connmanager.Notify(nodes[1:])
// Wait until connection manager replaces node connected in the first round.
require.NoError(t, utils.Eventually(func() error {
connected := server.Nodes()
if len(connected) != target {
return fmt.Errorf("unexpected number of connected servers: %d", len(connected))
}
switch enode.ID(connected[0]) {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
case nodes[1].ID():
case nodes[2].ID():
default:
return fmt.Errorf("connected with unexpected peer. got %s, expected %+v", connected[0], nodes[1:])
}
return nil
}, time.Second, 100*time.Millisecond))
}
func setupTestConnectionAfterExpiry(t *testing.T, server *fakePeerEvents, whisperMock *fakeEnvelopeEvents, target, maxFailures int, hash types.Hash) (*ConnectionManager, types.EnodeID) {
2019-01-21 14:00:10 +00:00
connmanager := NewConnectionManager(server, whisperMock, target, maxFailures, 0)
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
connmanager.Start()
nodes := []*enode.Node{}
for _, n := range getMapWithRandomNodes(t, 2) {
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
nodes = append(nodes, n)
}
// Send two random nodes to connection manager.
connmanager.Notify(nodes)
var initial types.EnodeID
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
// Wait until connection manager establishes connection with one node.
require.NoError(t, utils.Eventually(func() error {
nodes := server.Nodes()
if len(nodes) != target {
return fmt.Errorf("unexpected number of connected servers: %d", len(nodes))
}
initial = nodes[0]
return nil
}, time.Second, 100*time.Millisecond))
// Send event that history request for connected peer was sent.
select {
case whisperMock.input <- types.EnvelopeEvent{
Event: types.EventMailServerRequestSent, Peer: initial, Hash: hash}:
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
case <-time.After(time.Second):
require.FailNow(t, "can't send a 'sent' event")
}
2019-01-21 14:00:10 +00:00
return connmanager, initial
}
func TestConnectionChangedAfterExpiry(t *testing.T) {
server := newFakeServer()
whisperMock := newFakeEnvelopesEvents()
target := 1
maxFailures := 1
hash := types.Hash{1}
2019-01-21 14:00:10 +00:00
connmanager, initial := setupTestConnectionAfterExpiry(t, server, whisperMock, target, maxFailures, hash)
defer connmanager.Stop()
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
// And eventually expired.
select {
case whisperMock.input <- types.EnvelopeEvent{
Event: types.EventMailServerRequestExpired, Peer: initial, Hash: hash}:
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
case <-time.After(time.Second):
2019-01-21 14:00:10 +00:00
require.FailNow(t, "can't send an 'expiry' event")
}
require.NoError(t, utils.Eventually(func() error {
nodes := server.Nodes()
if len(nodes) != target {
return fmt.Errorf("unexpected number of connected servers: %d", len(nodes))
}
if nodes[0] == initial {
return fmt.Errorf("connected node wasn't changed from %s", initial)
}
return nil
}, time.Second, 100*time.Millisecond))
}
func TestConnectionChangedAfterSecondExpiry(t *testing.T) {
server := newFakeServer()
whisperMock := newFakeEnvelopesEvents()
target := 1
maxFailures := 2
hash := types.Hash{1}
2019-01-21 14:00:10 +00:00
connmanager, initial := setupTestConnectionAfterExpiry(t, server, whisperMock, target, maxFailures, hash)
defer connmanager.Stop()
// First expired is sent. Nothing should happen.
select {
case whisperMock.input <- types.EnvelopeEvent{
Event: types.EventMailServerRequestExpired, Peer: initial, Hash: hash}:
2019-01-21 14:00:10 +00:00
case <-time.After(time.Second):
require.FailNow(t, "can't send an 'expiry' event")
}
// we use 'eventually' as 'consistently' because this function will retry for a given timeout while error is received
require.EqualError(t, utils.Eventually(func() error {
nodes := server.Nodes()
if len(nodes) != target {
return fmt.Errorf("unexpected number of connected servers: %d", len(nodes))
}
if nodes[0] == initial {
return fmt.Errorf("connected node wasn't changed from %s", initial)
}
return nil
}, time.Second, 100*time.Millisecond), fmt.Sprintf("connected node wasn't changed from %s", initial))
// second expiry event
select {
case whisperMock.input <- types.EnvelopeEvent{
Event: types.EventMailServerRequestExpired, Peer: initial, Hash: hash}:
2019-01-21 14:00:10 +00:00
case <-time.After(time.Second):
Mail peer store and connection manager (#1295) This change implements connection manager that monitors 3 types of events: 1. update of the selected mail servers 2. disconnect from a mail server 3. errors for requesting mail history When selected mail servers provided we will try to connect with as many as possible, and later disconnect the surplus. For example if we want to connect with one mail server and 3 were selected, we try to connect with all (3), and later disconnect with 2. It will to establish connection with live mail server faster. If mail server disconnects we will choose any other mail server from the list of selected. Unless we have only one mail server. In such case we don't have any other choice and we will leave things as is. If request for history was expired we will disconnect such peer and try to find another one. We will follow same rules as described above. We will have two components that will rely on this logic: 1. requesting history If target peer is provided we will use that peer, otherwise we will request history from any selected mail server that is connected at the time of request. 2. confirmation from selected mail server Confirmation from any selected mail server will bee used to send a feedback that envelope was sent. I will add several extensions, but probably in separate PRs: 1. prioritize connection with mail server that was used before reboot 2. disconnect from mail servers if history request wasn't expired but failed. 3. wait some time in RequestsMessage RPC to establish connection with any mail server Currently this feature is hidden, as certain changes will be necessary in status-react. partially implements: https://github.com/status-im/status-go/issues/1285
2018-12-05 13:57:05 +00:00
require.FailNow(t, "can't send an 'expiry' event")
}
require.NoError(t, utils.Eventually(func() error {
nodes := server.Nodes()
if len(nodes) != target {
return fmt.Errorf("unexpected number of connected servers: %d", len(nodes))
}
if nodes[0] == initial {
return fmt.Errorf("connected node wasn't changed from %s", initial)
}
return nil
}, time.Second, 100*time.Millisecond))
}
func TestProcessReplacementWaitsForConnections(t *testing.T) {
srv := newFakePeerAdderRemover()
target := 1
timeout := time.Second
nodes := make([]*enode.Node, 2)
fillWithRandomNodes(t, nodes)
events := make(chan *p2p.PeerEvent)
state := newInternalState(srv, target, timeout)
state.currentNodes = nodesToMap(nodes)
go func() {
select {
case events <- &p2p.PeerEvent{Peer: nodes[0].ID(), Type: p2p.PeerEventTypeAdd}:
case <-time.After(time.Second):
assert.FailNow(t, "can't send a drop event")
}
}()
state.processReplacement(nodes, events)
require.Len(t, state.connected, 1)
}