1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
|
# OpenTelemetry
Requirements:
- docker-compose
## Prepare setup
Create a `docker-compose.yaml` and `otel-collector-config.yml` file as seen below in a folder.
```yaml title="docker-compose.yaml"
version: '3'
services:
# Jaeger
jaeger:
image: jaegertracing/all-in-one:1.64.0
ports:
- '16686:16686'
- '4317'
otel-collector:
image: otel/opentelemetry-collector-contrib:0.115.1
command: ['--config=/etc/otel-collector-config.yml']
volumes:
- ./otel-collector-config.yml:/etc/otel-collector-config.yml
ports:
- '1888:1888' # pprof extension
- '13133:13133' # health_check extension
- '55679:55679' # zpages extension
- '4318:4318' # OTLP HTTP
- '4317:4317' # OTLP GRPC
- '9123:9123' # Prometheus exporter
depends_on:
- jaeger
```
```yaml title="otel-collector-config.yml"
receivers:
otlp:
protocols:
grpc:
http:
exporters:
otlp/jaeger:
endpoint: jaeger:4317
tls:
insecure: true
logging:
prometheus:
endpoint: '0.0.0.0:9123'
processors:
batch:
spanmetrics:
metrics_exporter: prometheus
latency_histogram_buckets: [10ms, 100ms, 250ms, 1s, 30s, 1m, 5m]
dimensions:
- name: http.method
- name: http.status_code
- name: http.host
dimensions_cache_size: 1000
aggregation_temporality: 'AGGREGATION_TEMPORALITY_CUMULATIVE'
extensions:
health_check:
pprof:
zpages:
service:
extensions: [pprof, zpages, health_check]
pipelines:
traces:
receivers: [otlp]
exporters: [otlp/jaeger, logging]
processors: [spanmetrics, batch]
metrics:
receivers: [otlp]
exporters: [prometheus]
```
Start setup using this command inside the folder containing the files created in the earlier steps:
```
docker-compose up
```
This command will start an [OpenTelemetry Collector](https://github.com/open-telemetry/opentelemetry-collector-contrib) and an instance of [Jaeger](https://www.jaegertracing.io/).
Jaeger will be now reachable under [http://localhost:16686](http://localhost:16686).
## Run Renovate with OpenTelemetry
To start Renovate with OpenTelemetry enabled run following command, after pointing to your `config.js` config file:
```
docker run \
--rm \
-e OTEL_EXPORTER_OTLP_ENDPOINT=http://localhost:4318 \
-v "/path/to/your/config.js:/usr/src/app/config.js" \
renovate/renovate:latest
```
You should now see `trace_id` and `span_id` fields in the logs.
```
INFO: Repository finished (repository=org/example)
"durationMs": 5574,
"trace_id": "f9a4c33852333fc2a0fbdc163100c987",
"span_id": "4ac1323eeaee
```
### Traces
Open now Jaeger under [http://localhost:16686](http://localhost:16686).
You should now be able to pick `renovate` under in the field `service` field.
![service picker](../assets/images/opentelemetry_pick_service.png)
Select `Find Traces` to search for all Renovate traces and then select one of the found traces to open the trace view.
![pick trace](../assets/images/opentelemetry_choose_trace.png)
You should be able to see now the full trace view which shows each HTTP request and internal spans.
![trace view](../assets/images/opentelemetry_trace_viewer.png)
### Metrics
Additional to the received traces some metrics are calculated.
This is achieved using the [spanmetricsprocessor](https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/processor/spanmetricsprocessor).
The previous implemented setup will produce following metrics, which are exposed under [http://localhost:9123/metrics](http://localhost:9123/metrics):
```
# HELP calls_total
# TYPE calls_total counter
### Example of internal spans
calls_total{operation="renovate repository",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET"} 3
calls_total{operation="run",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET"} 1
### Example of http calls from Renovate to external services
calls_total{http_host="api.github.com:443",http_method="POST",http_status_code="200",operation="HTTPS POST",service_name="renovate",span_kind="SPAN_KIND_CLIENT",status_code="STATUS_CODE_UNSET"} 9
...
# HELP latency
# TYPE latency histogram
### Example of internal spans
latency_bucket{operation="renovate repository",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET",le="0.1"} 0
...
latency_bucket{operation="renovate repository",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET",le="9.223372036854775e+12"} 3
latency_bucket{operation="renovate repository",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET",le="+Inf"} 3
latency_sum{operation="renovate repository",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET"} 30947.4689
latency_count{operation="renovate repository",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET"} 3
...
### Example of http calls from Renovate to external services
latency_bucket{http_host="api.github.com:443",http_method="POST",http_status_code="200",operation="HTTPS POST",service_name="renovate",span_kind="SPAN_KIND_CLIENT",status_code="STATUS_CODE_UNSET",le="0.1"} 0
...
latency_bucket{http_host="api.github.com:443",http_method="POST",http_status_code="200",operation="HTTPS POST",service_name="renovate",span_kind="SPAN_KIND_CLIENT",status_code="STATUS_CODE_UNSET",le="250"} 3
latency_bucket{http_host="api.github.com:443",http_method="POST",http_status_code="200",operation="HTTPS POST",service_name="renovate",span_kind="SPAN_KIND_CLIENT",status_code="STATUS_CODE_UNSET",le="9.223372036854775e+12"} 9
latency_bucket{http_host="api.github.com:443",http_method="POST",http_status_code="200",operation="HTTPS POST",service_name="renovate",span_kind="SPAN_KIND_CLIENT",status_code="STATUS_CODE_UNSET",le="+Inf"} 9
latency_sum{http_host="api.github.com:443",http_method="POST",http_status_code="200",operation="HTTPS POST",service_name="renovate",span_kind="SPAN_KIND_CLIENT",status_code="STATUS_CODE_UNSET"} 2306.1385999999998
latency_count{http_host="api.github.com:443",http_method="POST",http_status_code="200",operation="HTTPS POST",service_name="renovate",span_kind="SPAN_KIND_CLIENT",status_code="STATUS_CODE_UNSET"} 9
```
The [spanmetricsprocessor](https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/processor/spanmetricsprocessor) creates two sets of metrics.
#### Calls metric
At first there are the `calls_total` metrics which display how often specific trace spans have been observed.
For example:
`calls_total{operation="renovate repository",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET"} 3` signals that 3 repositories have been renovated.
`calls_total{operation="run",service_name="renovate",span_kind="SPAN_KIND_INTERNAL",status_code="STATUS_CODE_UNSET"} 1` represents how often Renovate has been run.
If we combine this using the PrometheusQueryLanguage ( PromQL ), we can calculate the average count of repositories each Renovate run handles.
```
calls_total{operation="renovate repository",service_name="renovate"} / calls_total{operation="run",service_name="renovate"}
```
This metrics is also for spans generated by http calls:
```yaml
calls_total{http_host="registry.terraform.io:443",http_method="GET",http_status_code="200",operation="HTTPS GET",service_name="renovate",span_kind="SPAN_KIND_CLIENT",status_code="STATUS_CODE_UNSET"} 5
```
#### Latency buckets
The second class of metrics exposed are the latency focused latency buckets which allow to create [heatmaps](https://grafana.com/docs/grafana/latest/basics/intro-histograms/#heatmaps).
A request is added to a backed if the latency is bigger than the bucket value (`le`). `request_duration => le`
As an example if we receive a request which need `1.533s` to complete get following metrics:
```
latency_bucket{http_host="api.github.com:443",le="0.1"} 0
latency_bucket{http_host="api.github.com:443",le="1"} 0
latency_bucket{http_host="api.github.com:443",le="2"} 1
latency_bucket{http_host="api.github.com:443",le="6"} 1
latency_bucket{http_host="api.github.com:443",le="10"} 1
latency_bucket{http_host="api.github.com:443",le="100"} 1
latency_bucket{http_host="api.github.com:443",le="250"} 1
latency_bucket{http_host="api.github.com:443",le="9.223372036854775e+12"} 1
latency_bucket{http_host="api.github.com:443",le="+Inf"} 1
latency_sum{http_host="api.github.com:443"} 1.533
latency_count{http_host="api.github.com:443"} 1
```
Now we have another request which this time takes 10s to complete:
```
latency_bucket{http_host="api.github.com:443",le="0.1"} 0
latency_bucket{http_host="api.github.com:443",le="1"} 0
latency_bucket{http_host="api.github.com:443",le="2"} 1
latency_bucket{http_host="api.github.com:443",le="6"} 1
latency_bucket{http_host="api.github.com:443",le="10"} 2
latency_bucket{http_host="api.github.com:443",le="100"} 2
latency_bucket{http_host="api.github.com:443",le="250"} 2
latency_bucket{http_host="api.github.com:443",le="9.223372036854775e+12"} 2
latency_bucket{http_host="api.github.com:443",le="+Inf"} 2
latency_sum{http_host="api.github.com:443"} 11.533
latency_count{http_host="api.github.com:443"} 2
```
More about the functionality can be found on the Prometheus page for [metric types](https://prometheus.io/docs/concepts/metric_types/#histogram).
|