cmd/stdiscosrv: New discovery server (fixes #4618)
This is a new revision of the discovery server. Relevant changes and
non-changes:
- Protocol towards clients is unchanged.
- Recommended large scale design is still to be deployed nehind nginx (I
tested, and it's still a lot faster at terminating TLS).
- Database backend is leveldb again, only. It scales enough, is easy to
setup, and we don't need any backend to take care of.
- Server supports replication. This is a simple TCP channel - protect it
with a firewall when deploying over the internet. (We deploy this within
the same datacenter, and with firewall.) Any incoming client announces
are sent over the replication channel(s) to other peer discosrvs.
Incoming replication changes are applied to the database as if they came
from clients, but without the TLS/certificate overhead.
- Metrics are exposed using the prometheus library, when enabled.
- The database values and replication protocol is protobuf, because JSON
was quite CPU intensive when I tried that and benchmarked it.
- The "Retry-After" value for failed lookups gets slowly increased from
a default of 120 seconds, by 5 seconds for each failed lookup,
independently by each discosrv. This lowers the query load over time for
clients that are never seen. The Retry-After maxes out at 3600 after a
couple of weeks of this increase. The number of failed lookups is
stored in the database, now and then (avoiding making each lookup a
database put).
All in all this means clients can be pointed towards a cluster using
just multiple A / AAAA records to gain both load sharing and redundancy
(if one is down, clients will talk to the remaining ones).
GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/4648
2018-01-14 01:52:31 -07:00
|
|
|
// Copyright (C) 2018 The Syncthing Authors.
|
|
|
|
//
|
|
|
|
// This Source Code Form is subject to the terms of the Mozilla Public
|
|
|
|
// License, v. 2.0. If a copy of the MPL was not distributed with this file,
|
|
|
|
// You can obtain one at https://mozilla.org/MPL/2.0/.
|
2015-03-25 00:16:52 -07:00
|
|
|
|
|
|
|
package main
|
|
|
|
|
|
|
|
import (
|
2018-03-06 07:43:36 -07:00
|
|
|
"os"
|
|
|
|
|
cmd/stdiscosrv: New discovery server (fixes #4618)
This is a new revision of the discovery server. Relevant changes and
non-changes:
- Protocol towards clients is unchanged.
- Recommended large scale design is still to be deployed nehind nginx (I
tested, and it's still a lot faster at terminating TLS).
- Database backend is leveldb again, only. It scales enough, is easy to
setup, and we don't need any backend to take care of.
- Server supports replication. This is a simple TCP channel - protect it
with a firewall when deploying over the internet. (We deploy this within
the same datacenter, and with firewall.) Any incoming client announces
are sent over the replication channel(s) to other peer discosrvs.
Incoming replication changes are applied to the database as if they came
from clients, but without the TLS/certificate overhead.
- Metrics are exposed using the prometheus library, when enabled.
- The database values and replication protocol is protobuf, because JSON
was quite CPU intensive when I tried that and benchmarked it.
- The "Retry-After" value for failed lookups gets slowly increased from
a default of 120 seconds, by 5 seconds for each failed lookup,
independently by each discosrv. This lowers the query load over time for
clients that are never seen. The Retry-After maxes out at 3600 after a
couple of weeks of this increase. The number of failed lookups is
stored in the database, now and then (avoiding making each lookup a
database put).
All in all this means clients can be pointed towards a cluster using
just multiple A / AAAA records to gain both load sharing and redundancy
(if one is down, clients will talk to the remaining ones).
GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/4648
2018-01-14 01:52:31 -07:00
|
|
|
"github.com/prometheus/client_golang/prometheus"
|
2015-03-25 00:16:52 -07:00
|
|
|
)
|
|
|
|
|
cmd/stdiscosrv: New discovery server (fixes #4618)
This is a new revision of the discovery server. Relevant changes and
non-changes:
- Protocol towards clients is unchanged.
- Recommended large scale design is still to be deployed nehind nginx (I
tested, and it's still a lot faster at terminating TLS).
- Database backend is leveldb again, only. It scales enough, is easy to
setup, and we don't need any backend to take care of.
- Server supports replication. This is a simple TCP channel - protect it
with a firewall when deploying over the internet. (We deploy this within
the same datacenter, and with firewall.) Any incoming client announces
are sent over the replication channel(s) to other peer discosrvs.
Incoming replication changes are applied to the database as if they came
from clients, but without the TLS/certificate overhead.
- Metrics are exposed using the prometheus library, when enabled.
- The database values and replication protocol is protobuf, because JSON
was quite CPU intensive when I tried that and benchmarked it.
- The "Retry-After" value for failed lookups gets slowly increased from
a default of 120 seconds, by 5 seconds for each failed lookup,
independently by each discosrv. This lowers the query load over time for
clients that are never seen. The Retry-After maxes out at 3600 after a
couple of weeks of this increase. The number of failed lookups is
stored in the database, now and then (avoiding making each lookup a
database put).
All in all this means clients can be pointed towards a cluster using
just multiple A / AAAA records to gain both load sharing and redundancy
(if one is down, clients will talk to the remaining ones).
GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/4648
2018-01-14 01:52:31 -07:00
|
|
|
var (
|
|
|
|
apiRequestsTotal = prometheus.NewCounterVec(
|
|
|
|
prometheus.CounterOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "api_requests_total",
|
|
|
|
Help: "Number of API requests.",
|
|
|
|
}, []string{"type", "result"})
|
|
|
|
apiRequestsSeconds = prometheus.NewSummaryVec(
|
|
|
|
prometheus.SummaryOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "api_requests_seconds",
|
|
|
|
Help: "Latency of API requests.",
|
|
|
|
Objectives: map[float64]float64{0.5: 0.05, 0.9: 0.01, 0.99: 0.001},
|
|
|
|
}, []string{"type"})
|
|
|
|
|
|
|
|
lookupRequestsTotal = prometheus.NewCounterVec(
|
|
|
|
prometheus.CounterOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "lookup_requests_total",
|
|
|
|
Help: "Number of lookup requests.",
|
|
|
|
}, []string{"result"})
|
|
|
|
announceRequestsTotal = prometheus.NewCounterVec(
|
|
|
|
prometheus.CounterOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "announcement_requests_total",
|
|
|
|
Help: "Number of announcement requests.",
|
|
|
|
}, []string{"result"})
|
|
|
|
|
|
|
|
replicationSendsTotal = prometheus.NewCounterVec(
|
|
|
|
prometheus.CounterOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "replication_sends_total",
|
|
|
|
Help: "Number of replication sends.",
|
|
|
|
}, []string{"result"})
|
|
|
|
replicationRecvsTotal = prometheus.NewCounterVec(
|
|
|
|
prometheus.CounterOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "replication_recvs_total",
|
|
|
|
Help: "Number of replication receives.",
|
|
|
|
}, []string{"result"})
|
|
|
|
|
|
|
|
databaseKeys = prometheus.NewGaugeVec(
|
|
|
|
prometheus.GaugeOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "database_keys",
|
|
|
|
Help: "Number of database keys at last count.",
|
|
|
|
}, []string{"category"})
|
|
|
|
databaseStatisticsSeconds = prometheus.NewGauge(
|
|
|
|
prometheus.GaugeOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "database_statistics_seconds",
|
|
|
|
Help: "Time spent running the statistics routine.",
|
|
|
|
})
|
|
|
|
|
|
|
|
databaseOperations = prometheus.NewCounterVec(
|
|
|
|
prometheus.CounterOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "database_operations_total",
|
|
|
|
Help: "Number of database operations.",
|
|
|
|
}, []string{"operation", "result"})
|
|
|
|
databaseOperationSeconds = prometheus.NewSummaryVec(
|
|
|
|
prometheus.SummaryOpts{
|
|
|
|
Namespace: "syncthing",
|
|
|
|
Subsystem: "discovery",
|
|
|
|
Name: "database_operation_seconds",
|
|
|
|
Help: "Latency of database operations.",
|
|
|
|
Objectives: map[float64]float64{0.5: 0.05, 0.9: 0.01, 0.99: 0.001},
|
|
|
|
}, []string{"operation"})
|
|
|
|
)
|
2015-05-31 04:31:28 -07:00
|
|
|
|
cmd/stdiscosrv: New discovery server (fixes #4618)
This is a new revision of the discovery server. Relevant changes and
non-changes:
- Protocol towards clients is unchanged.
- Recommended large scale design is still to be deployed nehind nginx (I
tested, and it's still a lot faster at terminating TLS).
- Database backend is leveldb again, only. It scales enough, is easy to
setup, and we don't need any backend to take care of.
- Server supports replication. This is a simple TCP channel - protect it
with a firewall when deploying over the internet. (We deploy this within
the same datacenter, and with firewall.) Any incoming client announces
are sent over the replication channel(s) to other peer discosrvs.
Incoming replication changes are applied to the database as if they came
from clients, but without the TLS/certificate overhead.
- Metrics are exposed using the prometheus library, when enabled.
- The database values and replication protocol is protobuf, because JSON
was quite CPU intensive when I tried that and benchmarked it.
- The "Retry-After" value for failed lookups gets slowly increased from
a default of 120 seconds, by 5 seconds for each failed lookup,
independently by each discosrv. This lowers the query load over time for
clients that are never seen. The Retry-After maxes out at 3600 after a
couple of weeks of this increase. The number of failed lookups is
stored in the database, now and then (avoiding making each lookup a
database put).
All in all this means clients can be pointed towards a cluster using
just multiple A / AAAA records to gain both load sharing and redundancy
(if one is down, clients will talk to the remaining ones).
GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/4648
2018-01-14 01:52:31 -07:00
|
|
|
const (
|
|
|
|
dbOpGet = "get"
|
|
|
|
dbOpPut = "put"
|
|
|
|
dbOpMerge = "merge"
|
2018-05-16 00:26:20 -07:00
|
|
|
dbOpDelete = "delete"
|
cmd/stdiscosrv: New discovery server (fixes #4618)
This is a new revision of the discovery server. Relevant changes and
non-changes:
- Protocol towards clients is unchanged.
- Recommended large scale design is still to be deployed nehind nginx (I
tested, and it's still a lot faster at terminating TLS).
- Database backend is leveldb again, only. It scales enough, is easy to
setup, and we don't need any backend to take care of.
- Server supports replication. This is a simple TCP channel - protect it
with a firewall when deploying over the internet. (We deploy this within
the same datacenter, and with firewall.) Any incoming client announces
are sent over the replication channel(s) to other peer discosrvs.
Incoming replication changes are applied to the database as if they came
from clients, but without the TLS/certificate overhead.
- Metrics are exposed using the prometheus library, when enabled.
- The database values and replication protocol is protobuf, because JSON
was quite CPU intensive when I tried that and benchmarked it.
- The "Retry-After" value for failed lookups gets slowly increased from
a default of 120 seconds, by 5 seconds for each failed lookup,
independently by each discosrv. This lowers the query load over time for
clients that are never seen. The Retry-After maxes out at 3600 after a
couple of weeks of this increase. The number of failed lookups is
stored in the database, now and then (avoiding making each lookup a
database put).
All in all this means clients can be pointed towards a cluster using
just multiple A / AAAA records to gain both load sharing and redundancy
(if one is down, clients will talk to the remaining ones).
GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/4648
2018-01-14 01:52:31 -07:00
|
|
|
dbResSuccess = "success"
|
|
|
|
dbResNotFound = "not_found"
|
|
|
|
dbResError = "error"
|
|
|
|
dbResUnmarshalError = "unmarsh_err"
|
|
|
|
)
|
2015-05-31 04:31:28 -07:00
|
|
|
|
cmd/stdiscosrv: New discovery server (fixes #4618)
This is a new revision of the discovery server. Relevant changes and
non-changes:
- Protocol towards clients is unchanged.
- Recommended large scale design is still to be deployed nehind nginx (I
tested, and it's still a lot faster at terminating TLS).
- Database backend is leveldb again, only. It scales enough, is easy to
setup, and we don't need any backend to take care of.
- Server supports replication. This is a simple TCP channel - protect it
with a firewall when deploying over the internet. (We deploy this within
the same datacenter, and with firewall.) Any incoming client announces
are sent over the replication channel(s) to other peer discosrvs.
Incoming replication changes are applied to the database as if they came
from clients, but without the TLS/certificate overhead.
- Metrics are exposed using the prometheus library, when enabled.
- The database values and replication protocol is protobuf, because JSON
was quite CPU intensive when I tried that and benchmarked it.
- The "Retry-After" value for failed lookups gets slowly increased from
a default of 120 seconds, by 5 seconds for each failed lookup,
independently by each discosrv. This lowers the query load over time for
clients that are never seen. The Retry-After maxes out at 3600 after a
couple of weeks of this increase. The number of failed lookups is
stored in the database, now and then (avoiding making each lookup a
database put).
All in all this means clients can be pointed towards a cluster using
just multiple A / AAAA records to gain both load sharing and redundancy
(if one is down, clients will talk to the remaining ones).
GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/4648
2018-01-14 01:52:31 -07:00
|
|
|
func init() {
|
|
|
|
prometheus.MustRegister(apiRequestsTotal, apiRequestsSeconds,
|
|
|
|
lookupRequestsTotal, announceRequestsTotal,
|
|
|
|
replicationSendsTotal, replicationRecvsTotal,
|
|
|
|
databaseKeys, databaseStatisticsSeconds,
|
|
|
|
databaseOperations, databaseOperationSeconds)
|
2018-03-06 07:43:36 -07:00
|
|
|
|
2018-10-21 08:11:26 -07:00
|
|
|
processCollectorOpts := prometheus.ProcessCollectorOpts{
|
|
|
|
Namespace: "syncthing_discovery",
|
|
|
|
PidFn: func() (int, error) {
|
|
|
|
return os.Getpid(), nil
|
|
|
|
},
|
|
|
|
}
|
|
|
|
|
|
|
|
prometheus.MustRegister(
|
|
|
|
prometheus.NewProcessCollector(processCollectorOpts),
|
|
|
|
)
|
|
|
|
|
2015-05-31 04:31:28 -07:00
|
|
|
}
|