Merge branch 'develop' into units_in_router

This commit is contained in:
Thomas Gruber
2022-05-04 12:25:30 +02:00
committed by GitHub
35 changed files with 1497 additions and 541 deletions

View File

@@ -24,8 +24,8 @@ type ccMetric struct {
// ccMetric access functions
type CCMetric interface {
ToPoint(metaAsTags bool) *write.Point // Generate influxDB point for data type ccMetric
ToLineProtocol(metaAsTags bool) string // Generate influxDB line protocol for data type ccMetric
ToPoint(metaAsTags map[string]bool) *write.Point // Generate influxDB point for data type ccMetric
ToLineProtocol(metaAsTags map[string]bool) string // Generate influxDB line protocol for data type ccMetric
Name() string // Get metric name
SetName(name string) // Set metric name
@@ -61,25 +61,18 @@ func (m *ccMetric) String() string {
}
// ToLineProtocol generates influxDB line protocol for data type ccMetric
func (m *ccMetric) ToPoint(metaAsTags bool) (p *write.Point) {
if !metaAsTags {
p = influxdb2.NewPoint(m.name, m.tags, m.fields, m.tm)
} else {
tags := make(map[string]string, len(m.tags)+len(m.meta))
for key, value := range m.tags {
tags[key] = value
func (m *ccMetric) ToPoint(metaAsTags map[string]bool) (p *write.Point) {
p = influxdb2.NewPoint(m.name, m.tags, m.fields, m.tm)
for key, ok1 := range metaAsTags {
if val, ok2 := m.GetMeta(key); ok1 && ok2 {
p.AddTag(key, val)
}
for key, value := range m.meta {
tags[key] = value
}
p = influxdb2.NewPoint(m.name, tags, m.fields, m.tm)
}
return
return p
}
// ToLineProtocol generates influxDB line protocol for data type ccMetric
func (m *ccMetric) ToLineProtocol(metaAsTags bool) string {
func (m *ccMetric) ToLineProtocol(metaAsTags map[string]bool) string {
return write.PointToLineProtocol(
m.ToPoint(metaAsTags),

View File

@@ -169,7 +169,10 @@ func DieList() []int {
}
}
}
return dielist
if len(dielist) > 0 {
return dielist
}
return SocketList()
}
type CpuEntry struct {
@@ -261,7 +264,7 @@ func CpuData() []CpuEntry {
for _, c := range CpuList() {
clist = append(clist, CpuEntry{Cpuid: c})
}
for _, centry := range clist {
for i, centry := range clist {
centry.Socket = -1
centry.Numadomain = -1
centry.Die = -1
@@ -289,6 +292,8 @@ func CpuData() []CpuEntry {
// Lookup NUMA domain id
centry.Numadomain = getNumaDomain(base)
// Update values in output list
clist[i] = centry
}
return clist
}

View File

@@ -8,7 +8,8 @@ The CCMetric router sits in between the collectors and the sinks and can be used
{
"num_cache_intervals" : 1,
"interval_timestamp" : true,
"normalize_units": true,
"hostname_tag" : "hostname",
"max_forward" : 50,
"add_tags" : [
{
"key" : "cluster",
@@ -61,6 +62,20 @@ The CCMetric router sits in between the collectors and the sinks and can be used
```
There are three main options `add_tags`, `delete_tags` and `interval_timestamp`. `add_tags` and `delete_tags` are lists consisting of dicts with `key`, `value` and `if`. The `value` can be omitted in the `delete_tags` part as it only uses the `key` for removal. The `interval_timestamp` setting means that a unique timestamp is applied to all metrics traversing the router during an interval.
# Processing order in the router
- Add the `hostname_tag` tag (if sent by collectors or cache)
- If `interval_timestamp == true`, change time of metrics
- Check if metric should be dropped (`drop_metrics` and `drop_metrics_if`)
- Add tags from `add_tags`
- Delete tags from `del_tags`
- Rename metric based on `rename_metrics` and store old name as `oldname` in meta information
- Add tags from `add_tags` (if you used the new name in the `if` condition)
- Delete tags from `del_tags` (if you used the new name in the `if` condition)
- Send to sinks
- Move to cache (if `num_cache_intervals > 0`)
# The `interval_timestamp` option
The collectors' `Read()` functions are not called simultaneously and therefore the metrics gathered in an interval can have different timestamps. If you want to avoid that and have a common timestamp (the beginning of the interval), set this option to `true` and the MetricRouter sets the time.
@@ -71,6 +86,14 @@ If the MetricRouter should buffer metrics of intervals in a MetricCache, this op
A `num_cache_intervals > 0` is required to use the `interval_aggregates` option.
# The `hostname_tag` option
By default, the router tags metrics with the hostname for all locally created metrics. The default tag name is `hostname`, but it can be changed if your organization wants anything else
# The `max_forward` option
Every time the router receives a metric through any of the channels, it tries to directly read up to `max_forward` metrics from the same channel. This was done as the router thread would go to sleep and wake up with every arriving metric. The default are `50` metrics at once and `max_forward` needs to greater than `1`.
# The `rename_metrics` option
In the ClusterCockpit world we specified a set of standard metrics. Since some collectors determine the metric names based on files, execuables and libraries, they might change from system to system (or installation to installtion, OS to OS, ...). In order to get the common names, you can rename incoming metrics before sending them to the sink. If the metric name matches the `oldname`, it is changed to `newname`

View File

@@ -51,7 +51,6 @@ type metricRouter struct {
done chan bool // channel to finish / stop metric router
wg *sync.WaitGroup // wait group for all goroutines in cc-metric-collector
timestamp time.Time // timestamp periodically updated by ticker each interval
timerdone chan bool // channel to finish / stop timestamp updater
ticker mct.MultiChanTicker // periodically ticking once each interval
config metricRouterConfig // json encoded config for metric router
cache MetricCache // pointer to MetricCache
@@ -106,7 +105,10 @@ func (r *metricRouter) Init(ticker mct.MultiChanTicker, wg *sync.WaitGroup, rout
cclog.ComponentError("MetricRouter", err.Error())
return err
}
r.maxForward = r.config.MaxForward
r.maxForward = 1
if r.config.MaxForward > r.maxForward {
r.maxForward = r.config.MaxForward
}
if r.config.NumCacheIntervals > 0 {
r.cache, err = NewCache(r.cache_input, r.ticker, &r.cachewg, r.config.NumCacheIntervals)
if err != nil {
@@ -124,29 +126,6 @@ func (r *metricRouter) Init(ticker mct.MultiChanTicker, wg *sync.WaitGroup, rout
return nil
}
// StartTimer starts a timer which updates timestamp periodically
func (r *metricRouter) StartTimer() {
m := make(chan time.Time)
r.ticker.AddChannel(m)
r.timerdone = make(chan bool)
r.wg.Add(1)
go func() {
defer r.wg.Done()
for {
select {
case <-r.timerdone:
close(r.timerdone)
cclog.ComponentDebug("MetricRouter", "TIMER DONE")
return
case t := <-m:
r.timestamp = t
}
}
}()
cclog.ComponentDebug("MetricRouter", "TIMER START")
}
func getParamMap(point lp.CCMetric) map[string]interface{} {
params := make(map[string]interface{})
params["metric"] = point
@@ -267,8 +246,9 @@ func (r *metricRouter) prepareUnit(point lp.CCMetric) bool {
func (r *metricRouter) Start() {
// start timer if configured
r.timestamp = time.Now()
timeChan := make(chan time.Time)
if r.config.IntervalStamp {
r.StartTimer()
r.ticker.AddChannel(timeChan)
}
// Router manager is done
@@ -350,6 +330,10 @@ func (r *metricRouter) Start() {
done()
return
case timestamp := <-timeChan:
r.timestamp = timestamp
cclog.ComponentDebug("MetricRouter", "Update timestamp", r.timestamp.UnixNano())
case p := <-r.coll_input:
coll_forward(p)
for i := 0; len(r.coll_input) > 0 && i < (r.maxForward-1); i++ {
@@ -395,14 +379,6 @@ func (r *metricRouter) Close() {
// wait for close of channel r.done
<-r.done
// stop timer
if r.config.IntervalStamp {
cclog.ComponentDebug("MetricRouter", "TIMER CLOSE")
r.timerdone <- true
// wait for close of channel r.timerdone
<-r.timerdone
}
// stop metric cache
if r.config.NumCacheIntervals > 0 {
cclog.ComponentDebug("MetricRouter", "CACHE CLOSE")