Files
golib/nutsdb/model.go
nabbar e3239db998 Rework Monitoring, Prometheus, Status system
Package Monitoring :
- use packag dedicated to monitor component
- each monitor work as standalone server to monitor health
- collect metrics to export them to prometheus exporter

Package Prometheus :
- review to simplify use for API and not API metrics
- optimize code

Package Status :
- Rework to use Monitor package
- Rework to use native json / text Marshaller interface

Context :
- rework context config (context var) to use sync map and sync RWMutex (WORM)
- move gin context to dedicated sub package (dependancies of logger make circular dependencies)
- optimize code

Config :
- rework to optimize sync / collect of component
- rework status to monitor
- remove monitor managment from config to each component
- add a func to set default logger to implement inherit default logger options
- optimize code

IOUtils :
- isolate logger / closer interface as a usable & public interface & instance
- this interface / instance allow to collect io.closer over a context to close all if context is done

Logger :
- rework to use context.config map
- rework to use ioutils closer
- rework to allow options to inherit a default options, or the last version of options
- optimize code

Size :
- Add package Size to calculate and manipulate size Byte or bit
- Add encoding : Text/JSON/Yaml/Toml...
- Add option to défine default unit : Byte or bit

Other :
- adjust following code
- optimize code
- limit use of atomic value
- rework to use RWMutex instead of sync.Mutex to maximize capabilities of read instead of write
- remove 32bit build for CI/CD
- add darwin/arm64 build for CI/CD

Bump Dependencies
2023-02-21 16:30:40 +01:00

348 lines
8.4 KiB
Go

//go:build !386 && !arm && !mips && !mipsle
// +build !386,!arm,!mips,!mipsle
/***********************************************************************************************************************
*
* MIT License
*
* Copyright (c) 2021 Nicolas JUHEL
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in all
* copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
*
*
**********************************************************************************************************************/
package nutsdb
import (
"context"
"sync"
"sync/atomic"
"time"
dgbstm "github.com/lni/dragonboat/v3/statemachine"
libclu "github.com/nabbar/golib/cluster"
liberr "github.com/nabbar/golib/errors"
liblog "github.com/nabbar/golib/logger"
libsh "github.com/nabbar/golib/shell"
)
type ndb struct {
m sync.Mutex
c Config
l liblog.FuncLog
e liberr.Error
t *atomic.Value // cluster
r *atomic.Value // status
}
func (n *ndb) createNodeMachine(node uint64, cluster uint64) dgbstm.IOnDiskStateMachine {
var (
err liberr.Error
opt Options
)
if opt, err = n.c.GetOptions(); err != nil {
panic(err)
}
return newNode(node, cluster, opt, n.setRunning)
}
func (n *ndb) newCluster() liberr.Error {
var (
clu libclu.Cluster
err liberr.Error
cfg libclu.Config
)
if c := n.Cluster(); c != nil {
if err = n.Shutdown(); err != nil {
return err
}
}
if cfg, err = n.c.GetConfigCluster(); err != nil {
return err
}
clu, err = libclu.NewCluster(cfg, nil)
if err != nil {
return err
}
clu.SetFctCreateSTMOnDisk(n.createNodeMachine)
n.setCluster(clu)
return nil
}
func (n *ndb) GetLogger() liblog.Logger {
n.m.Lock()
defer n.m.Unlock()
if n.l != nil {
return n.l()
}
return liblog.GetDefault()
}
func (n *ndb) SetLogger(l liblog.FuncLog) {
n.m.Lock()
defer n.m.Unlock()
n.l = l
}
func (n *ndb) IsRunning() bool {
n.m.Lock()
defer n.m.Unlock()
if i := n.r.Load(); i == nil {
return false
} else if b, ok := i.(bool); !ok {
return false
} else {
return b
}
}
func (n *ndb) setRunning(state bool) {
n.m.Lock()
defer n.m.Unlock()
if n == nil || n.r == nil {
return
} else {
n.r.Store(state)
}
}
func (n *ndb) IsReady(ctx context.Context) bool {
if m, e := n.Cluster().SyncGetClusterMembership(ctx); e != nil || m == nil || len(m.Nodes) < 1 {
return false
}
if _, ok, e := n.Cluster().GetLeaderID(); e != nil || !ok {
return false
} else {
return true
}
}
func (n *ndb) IsReadyTimeout(parent context.Context, dur time.Duration) bool {
ctx, cnl := context.WithTimeout(parent, dur)
defer cnl()
if n.IsRunning() && n.IsReady(ctx) {
return true
}
return false
}
func (n *ndb) WaitReady(ctx context.Context, tick time.Duration) {
for {
if n.IsRunning() && n.IsReady(ctx) {
return
}
time.Sleep(tick)
}
}
func (n *ndb) Listen() liberr.Error {
var (
c libclu.Cluster
e liberr.Error
)
if c = n.Cluster(); c == nil {
if e = n.newCluster(); e != nil {
n._SetError(e)
return e
} else if c = n.Cluster(); c == nil {
n._SetError(e)
return ErrorClusterInit.Error(nil)
}
}
if e = c.ClusterStart(len(n.c.Cluster.InitMember) < 1); e != nil {
n._SetError(e)
return e
}
n._SetError(nil)
n.setCluster(c)
return nil
}
func (n *ndb) Restart() liberr.Error {
return n.Listen()
}
func (n *ndb) ForceRestart() {
n.ForceShutdown()
_ = n.Listen()
}
func (n *ndb) Shutdown() liberr.Error {
if c := n.Cluster(); c == nil {
return nil
} else if !c.HasNodeInfo(0) {
return nil
} else if err := c.NodeStop(0); err != nil {
return err
} else {
n.setCluster(c)
return nil
}
}
func (n *ndb) ForceShutdown() {
if err := n.Shutdown(); err == nil {
return
} else if c := n.Cluster(); c == nil {
return
} else if !c.HasNodeInfo(0) {
return
} else {
_ = c.ClusterStop(true)
n.setCluster(c)
}
}
func (n *ndb) Cluster() libclu.Cluster {
n.m.Lock()
defer n.m.Unlock()
if i := n.t.Load(); i == nil {
return nil
} else if c, ok := i.(libclu.Cluster); !ok {
return nil
} else {
return c
}
}
func (n *ndb) setCluster(clu libclu.Cluster) {
n.m.Lock()
defer n.m.Unlock()
n.t.Store(clu)
}
func (n *ndb) Client(ctx context.Context, tickSync time.Duration) Client {
return &clientNutDB{
x: ctx,
t: tickSync,
c: n.Cluster,
w: n.WaitReady,
}
}
func (n *ndb) ShellCommand(ctx func() context.Context, tickSync time.Duration) []libsh.Command {
var (
res = make([]libsh.Command, 0)
cli func() Client
)
cli = func() Client {
x := ctx()
if x.Err() != nil {
return nil
}
return n.Client(x, tickSync)
}
res = append(res, newShellCommand(CmdPut, cli))
res = append(res, newShellCommand(CmdPutWithTimestamp, cli))
res = append(res, newShellCommand(CmdGet, cli))
res = append(res, newShellCommand(CmdGetAll, cli))
res = append(res, newShellCommand(CmdRangeScan, cli))
res = append(res, newShellCommand(CmdPrefixScan, cli))
res = append(res, newShellCommand(CmdPrefixSearchScan, cli))
res = append(res, newShellCommand(CmdDelete, cli))
res = append(res, newShellCommand(CmdFindTxIDOnDisk, cli))
res = append(res, newShellCommand(CmdFindOnDisk, cli))
res = append(res, newShellCommand(CmdFindLeafOnDisk, cli))
res = append(res, newShellCommand(CmdSAdd, cli))
res = append(res, newShellCommand(CmdSRem, cli))
res = append(res, newShellCommand(CmdSAreMembers, cli))
res = append(res, newShellCommand(CmdSIsMember, cli))
res = append(res, newShellCommand(CmdSMembers, cli))
res = append(res, newShellCommand(CmdSHasKey, cli))
res = append(res, newShellCommand(CmdSPop, cli))
res = append(res, newShellCommand(CmdSCard, cli))
res = append(res, newShellCommand(CmdSDiffByOneBucket, cli))
res = append(res, newShellCommand(CmdSDiffByTwoBuckets, cli))
res = append(res, newShellCommand(CmdSMoveByOneBucket, cli))
res = append(res, newShellCommand(CmdSMoveByTwoBuckets, cli))
res = append(res, newShellCommand(CmdSUnionByOneBucket, cli))
res = append(res, newShellCommand(CmdSUnionByTwoBuckets, cli))
res = append(res, newShellCommand(CmdRPop, cli))
res = append(res, newShellCommand(CmdRPeek, cli))
res = append(res, newShellCommand(CmdRPush, cli))
res = append(res, newShellCommand(CmdLPush, cli))
res = append(res, newShellCommand(CmdLPop, cli))
res = append(res, newShellCommand(CmdLPeek, cli))
res = append(res, newShellCommand(CmdLSize, cli))
res = append(res, newShellCommand(CmdLRange, cli))
res = append(res, newShellCommand(CmdLRem, cli))
res = append(res, newShellCommand(CmdLSet, cli))
res = append(res, newShellCommand(CmdLTrim, cli))
res = append(res, newShellCommand(CmdZAdd, cli))
res = append(res, newShellCommand(CmdZMembers, cli))
res = append(res, newShellCommand(CmdZCard, cli))
res = append(res, newShellCommand(CmdZCount, cli))
res = append(res, newShellCommand(CmdZPopMax, cli))
res = append(res, newShellCommand(CmdZPopMin, cli))
res = append(res, newShellCommand(CmdZPeekMax, cli))
res = append(res, newShellCommand(CmdZPeekMin, cli))
res = append(res, newShellCommand(CmdZRangeByScore, cli))
res = append(res, newShellCommand(CmdZRangeByRank, cli))
res = append(res, newShellCommand(CmdZRem, cli))
res = append(res, newShellCommand(CmdZRemRangeByRank, cli))
res = append(res, newShellCommand(CmdZRank, cli))
res = append(res, newShellCommand(CmdZRevRank, cli))
res = append(res, newShellCommand(CmdZScore, cli))
res = append(res, newShellCommand(CmdZGetByKey, cli))
return res
}
func (n *ndb) _SetError(e liberr.Error) {
n.m.Lock()
defer n.m.Unlock()
n.e = e
}
func (n *ndb) _GetError() liberr.Error {
n.m.Lock()
defer n.m.Unlock()
return n.e
}