Skip to content

Commit

Permalink
docs(sql-exporter): updated deployments and architecture
Browse files Browse the repository at this point in the history
  • Loading branch information
qfritz committed Feb 2, 2024
1 parent 1328088 commit a8ca611
Show file tree
Hide file tree
Showing 4 changed files with 53 additions and 1,222 deletions.
17 changes: 5 additions & 12 deletions content/architecture.md
Original file line number Diff line number Diff line change
Expand Up @@ -17,24 +17,17 @@ The [Prometheus RDS exporter](https://github.com/qonto/prometheus-rds-exporter)

This component is optional but highly recommended for AWS customers.

### Prometheus PostgreSQL exporter
### PostgreSQL exporter

{{< hint info >}}
**Tips**
A generic [SQL exporter](https://github.com/burningalchemist/sql_exporter) collects PostgreSQL internal metrics using system tables.

This section is outdated for DMF version >beta.10 because of ongoing migration toward burningalchemist/sql_exporter. Stay tuned!

{{< /hint >}}

The [Prometheus Postgres exporter](https://github.com/prometheus-community/postgres_exporter) collects PostgreSQL internal metrics using system tables.

We override its default configuration to collect additional metrics to trigger advanced alerts and analyze trends over time.
We have our own set of collectors, retrieving metrics to trigger advanced alerts and analyze trends over time. All our collectors for PostgreSQL can be found in [postgresql-collectors](https://github.com/qonto/postgresql-collectors).

This component is optional but highly recommended to analyze PostgreSQL metrics over time.

### Prometheus alerts

Monitoring alerts are defined in Prometheus rule format and work out of the box with the RDS and PostgreSQL exporters in Kubernetes environment with Prometheus operator.
Monitoring alerts are defined in Prometheus rule format and work out of the box with the RDS and SQL exporters in Kubernetes environment with Prometheus operator.

They are distributed via Helm chart for convenient deployment:

Expand All @@ -43,7 +36,7 @@ They are distributed via Helm chart for convenient deployment:
| [prometheus-rds-alerts-chart](https://gallery.ecr.aws/qonto/prometheus-rds-alerts-chart) | Alerts for AWS RDS |
| [prometheus-postgresql-alerts-chart](https://gallery.ecr.aws/qonto/prometheus-postgresql-alerts-chart) | Alerts for PostgreSQL |

The charts deploy [Prometheus `PrometheusRule`](https://github.com/prometheus-operator/prometheus-operator/blob/main/Documentation/user-guides/alerting.md#deploying-prometheus-rules) Kubernetes resources. So Prometheus automatically discover the alerts.
These charts deploy [Prometheus `PrometheusRule`](https://github.com/prometheus-operator/prometheus-operator/blob/main/Documentation/user-guides/alerting.md#deploying-prometheus-rules) Kubernetes resources. So Prometheus automatically discovers the alerts.

### Runbooks

Expand Down
Loading

0 comments on commit a8ca611

Please sign in to comment.