Graph Sink¶
Overview¶
The Unified Assurance Graph Sink microservice is part of the microservice topology pipeline. The Graph Sink is responsible for topology data that has been normalized. It and inserts the normalized topology data into the Neo4j database.
Prerequisites¶
-
A microservices cluster must be setup. Refer to Microservice Cluster Setup.
-
Apache Pulsar must be installed. Refer to Apache Pulsar microservice.
Setup¶
su - assure1
export NAMESPACE=a1-zone1-pri
export WEBFQDN=<Primary Presentation Web FQDN>
a1helm install graph-sink assure1/graph-sink -n $NAMESPACE --set global.imageRegistry=$WEBFQDN
Default Configuration¶
Name | Value | Possible Values | Notes |
---|---|---|---|
LOG_LEVEL | INFO | FATAL, ERROR, WARN, INFO, DEBUG | Logging level used by application. |
STREAM_INPUT | persistent://assure1/graph/sink | Text, 255 characters | Apache Pulsar topic path. Topic at end of path may be any text value. |
BATCH_DELAY | 0 | Integer | Time to wait in seconds before processing a batch. If set to 0, messages will be processed one-by-one. |
MAX_BATCH_SIZE | 1000 | Integer | Maximum number of messages that can be present in a batch. |
Configurations can be changed by passing the values to the a1helm install
prefixed with the configData parent key.
Example of setting the log level to DEBUG¶
a1helm install ... --set configData.LOG_LEVEL=DEBUG
Autoscaling¶
For microservice scaling options, please refer to the autoscaling docs.
On top of the standard autoscaling options, this microservice supports the following additional configurations:
Name | Value | Possible Values | Notes |
---|---|---|---|
thresholds.backlogSize | 1000 | Integer | The number of items that need to be in the backlog before the auto-scaling starts additional processes. |
thresholds.totalItemsProcessed | 400 | Integer | Total items processed by the microservice. If the average of total items processed in five minutes exceeds the threshold, pods will be scaled. |
Example of changing the auto-scaling value¶
To change the thresholds used for this functionality, pass the following to the installation command:
a1helm install ... --set autoscaling.thresholds.backlogSize=2000
Schema¶
Below is an example of JSON file formats with the accepted schema.
Single Vertex Schema
{
"@timestamp": "2020-03-18T18:41:00.000Z",
"_agents": [
{
"@timestamp": "2020-03-18T18:40:00.000Z",
"host": "host.example.com",
"id": "d5a346c8-d675-48a1-a42f-4a8db7c1552b",
"app": "sample-integration",
"version": "5.0.0",
"input": "unknown://",
"output": "pulsar+ssl://proxy.a1-messaging.svc.cluster.local/assure1/graph/sink"
}
],
"_domain": "topology",
"_type": "vertex",
"_version": "5.0.0",
"vertex": {
"@meta": false,
"@method": "upsert",
"@type": "Device",
"keys": {
"Name": "device-a.example.com",
"ZoneID": 1
},
"properties": {
"TimestampModified": "123456789000"
}
}
}
Edge Tuple Schema
{
"@timestamp": "2021-03-30T15:32:21.690779663Z",
"_agents": [
{
"@timestamp": "2021-03-30T15:32:21.690778749Z",
"host": "pg-example.example.com",
"id": "255c83bd-ff1e-2e10-6d92-dfecce7fba21",
"app": "virtuora-connector",
"version": "1.0.1",
"input": "pg-example.example.com",
"output": "persistent://assure1/graph/sink"
}
],
"_domain": "topology",
"_type": "edgeTuple",
"_version": "1.0.1",
"edgeTuple": {
"outVertex": {
"@meta": false,
"@method": "upsert",
"@type": "Device",
"keys": {
"Name": "L100-107-TEST",
"ZoneID": 1
},
"properties": {
"TimestampModified": 1617118341
}
},
"inVertex": {
"@meta": false,
"@method": "upsert",
"@type": "Device",
"keys": {
"Name": "L100-107-TEST",
"ZoneID": 0
},
"properties": {
"TimestampModified": 1617118341
}
},
"edge": {
"@meta": false,
"@method": "upsert",
"@type": "ConnectsInterface",
"keys": null,
"properties": {
"TimestampModified": 1617118341
}
}
}
}
Deletion by single vertex example, type and at least one key is required:
{
"@timestamp": "2020-03-18T18:41:00.000Z",
"_agents": [
{
"@timestamp": "2020-03-18T18:40:00.000Z",
"host": "host.example.com",
"id": "d5a346c8-d675-48a1-a42f-4a8db7c1552b",
"app": "sample-integration",
"version": "5.0.0",
"input": "unknown://",
"output": "pulsar+ssl://proxy.a1-messaging.svc.cluster.local/assure1/graph/sink"
}
],
"_domain": "topology",
"_type": "vertex",
"_version": "5.0.0",
"vertex": {
"@meta": false,
"@method": "delete",
"@type": "Device",
"keys": {
"Name": "device-x.example.com",
"ZoneID": 1
}
}
}
Deletion by single edge, type and at least one unique property is required:
{
"@timestamp": "2020-03-18T18:41:00.000Z",
"_agents": [
{
"@timestamp": "2020-03-18T18:40:00.000Z",
"host": "host.example.com",
"id": "d5a346c8-d675-48a1-a42f-4a8db7c1552b",
"app": "sample-integration",
"version": "5.0.0",
"input": "unknown://",
"output": "pulsar+ssl://proxy.a1-messaging.svc.cluster.local/assure1/graph/sink"
}
],
"_domain": "topology",
"_type": "edge",
"_version": "5.0.0",
"edge": {
"@meta": false,
"@method": "delete",
"@type": "HasInterface",
"properties": {
"connectionID": "1023"
}
}
}
Microservice self-metrics¶
The Unified Assurance Graph Sink microservice exposes the following self-metrics to Prometheus.
Metric Name | Type | Description |
---|---|---|
graph_sink_backlog_size | Gauge | Number of items pending in the backlog |
total_items_processed | Counter | Number of items processed |
processing_time_per_item | Gauge | Processing time per item from receiving it to sending it to acknowledging it |