Bläddra i källkod

Merge f69b61aaf0 into 7029f48322

pull/495/merge
gabriel-tessier 1 dag sedan
committed by GitHub
förälder
incheckning
a192fe9d58
Ingen känd nyckel hittad för denna signaturen i databasen GPG-nyckel ID: B5690EEEBB952194
3 ändrade filer med 127 tillägg och 1 borttagningar
  1. +127
    -1
      docs/guides/edge.md
  2. Binär
      website/static/img/advanced_web_service_with_on-premise_less_edges.png
  3. Binär
      website/static/img/advanced_web_service_with_on-premise_merged_edges.png

+ 127
- 1
docs/guides/edge.md Visa fil

@@ -65,5 +65,131 @@ with Diagram(name="Advanced Web Service with On-Premises (colored)", show=False)
>> Edge(color="darkorange") \
>> aggregator
```
![advanced web service with on-premise diagram colored](/img/advanced_web_service_with_on-premise_colored.png)

![advanced web service with on-premises diagram colored](/img/advanced_web_service_with_on-premises_colored.png)
## Less Edges

As you can see on the previous graph the edges can quickly become noisy. Below are two examples to solve this problem.

One approach is to get creative with the Node class to create blank placeholders, together with named nodes within Clusters, and then only pointing to single named elements within those Clusters.

Compare the output below to the example output above .

```python
from diagrams import Cluster, Diagram, Node
from diagrams.onprem.analytics import Spark
from diagrams.onprem.compute import Server
from diagrams.onprem.database import PostgreSQL
from diagrams.onprem.inmemory import Redis
from diagrams.onprem.aggregator import Fluentd
from diagrams.onprem.monitoring import Grafana, Prometheus
from diagrams.onprem.network import Nginx
from diagrams.onprem.queue import Kafka

with Diagram("\nAdvanced Web Service with On-Premise Less edges", show=False) as diag:
ingress = Nginx("ingress")

with Cluster("Service Cluster"):
serv1 = Server("grpc1")
serv2 = Server("grpc2")
serv3 = Server("grpc3")

with Cluster(""):
blankHA = Node("", shape="plaintext", width="0", height="0")

metrics = Prometheus("metric")
metrics << Grafana("monitoring")

aggregator = Fluentd("logging")
blankHA >> aggregator >> Kafka("stream") >> Spark("analytics")

with Cluster("Database HA"):
db = PostgreSQL("users")
db - PostgreSQL("replica") << metrics
blankHA >> db

with Cluster("Sessions HA"):
sess = Redis("session")
sess - Redis("replica") << metrics
blankHA >> sess

ingress >> serv2 >> blankHA

diag
```

![advanced web service with on-premise less edges](/img/advanced_web_service_with_on-premise_less_edges.png)

## Merged Edges

Yet another option is to set the graph_attr dictionary key "concentrate" to "true".

Note the following restrictions:

1. the Edge must end at the same headport
2. This only works when the "splines" graph_attr key is set to the value "spline". It has no effect when the value was set to "ortho", which is the default for the diagrams library.
3. this will only work with the "dot" layout engine, which is the default for the diagrams library.

For more information see:

https://graphviz.gitlab.io/doc/info/attrs.html#d:concentrate

https://www.graphviz.org/pdf/dotguide.pdf Section 3.3 Concentrators



```python
from diagrams import Cluster, Diagram, Edge, Node
from diagrams.onprem.analytics import Spark
from diagrams.onprem.compute import Server
from diagrams.onprem.database import PostgreSQL
from diagrams.onprem.inmemory import Redis
from diagrams.onprem.aggregator import Fluentd
from diagrams.onprem.monitoring import Grafana, Prometheus
from diagrams.onprem.network import Nginx
from diagrams.onprem.queue import Kafka

graph_attr = {
"concentrate": "true",
"splines": "spline",
}

edge_attr = {
"minlen":"3",
}

with Diagram("\n\nAdvanced Web Service with On-Premise Merged edges", show=False,
graph_attr=graph_attr,
edge_attr=edge_attr) as diag:

ingress = Nginx("ingress")

metrics = Prometheus("metric")
metrics << Edge(minlen="0") << Grafana("monitoring")

with Cluster("Service Cluster"):
grpsrv = [
Server("grpc1"),
Server("grpc2"),
Server("grpc3")]

blank = Node("", shape="plaintext", height="0.0", width="0.0")

with Cluster("Sessions HA"):
sess = Redis("session")
sess - Redis("replica") << metrics

with Cluster("Database HA"):
db = PostgreSQL("users")
db - PostgreSQL("replica") << metrics

aggregator = Fluentd("logging")
aggregator >> Kafka("stream") >> Spark("analytics")

ingress >> [grpsrv[0], grpsrv[1], grpsrv[2],]
[grpsrv[0], grpsrv[1], grpsrv[2],] - Edge(headport="w", minlen="1") - blank
blank >> Edge(headport="w", minlen="2") >> [sess, db, aggregator]

diag
```
![advanced web service with on-premise merged edges](/img/advanced_web_service_with_on-premise_merged_edges.png)

Binär
website/static/img/advanced_web_service_with_on-premise_less_edges.png Visa fil

Före Efter
Bredd: 1647  |  Höjd: 1191  |  Storlek: 156 KiB

Binär
website/static/img/advanced_web_service_with_on-premise_merged_edges.png Visa fil

Före Efter
Bredd: 1932  |  Höjd: 1204  |  Storlek: 158 KiB

Laddar…
Avbryt
Spara