--- title: "Monitor Aerospike with Datadog" description: "Configure Datadog to collect and visualize Aerospike Enterprise metrics from Aerospike Prometheus Exporter." --- # Monitor Aerospike with Datadog > For the complete documentation index see: [llms.txt](https://aerospike.com/docs/llms.txt) > > All documentation pages available in markdown. This page shows how to configure Datadog to collect Aerospike Database Enterprise Edition metrics from Aerospike Prometheus Exporter. When you finish, Aerospike metrics are available in Datadog through either the Datadog Agent integration or a direct OpenTelemetry (OTel) HTTP connection. Use Datadog monitoring when your operations, SRE, or DevOps team already uses Datadog for infrastructure and application monitoring. Datadog can then collect Aerospike metrics for cluster availability, namespaces, nodes, sets, secondary indexes, XDR, and latency metrics. ## Choose a Datadog path You can connect Aerospike Database to Datadog through the Datadog Agent integration or directly with OTel HTTP. ### Datadog Agent Datadog Agent runs on the systems you monitor and collects metrics locally. The Aerospike Enterprise Agent integration uses the OpenMetrics endpoint exposed by Aerospike Prometheus Exporter, then sends the collected metrics to Datadog with the rest of your infrastructure telemetry. Use this path when you already run Datadog Agent or want Agent status, host metadata, Agent-level tagging, and standard integration troubleshooting. ### Direct OTel HTTP If you prefer to use direct OTel HTTP, Aerospike Prometheus Exporter sends OTel metrics directly to the Datadog OTLP metrics intake endpoint, so no Datadog Agent or OTel Collector is required between the exporter and Datadog. Use this path when you cannot run Datadog Agent near the exporter, prefer outbound HTTPS from the exporter, or only need to send Aerospike metrics. This requires the exporter to store a Datadog API key and have outbound access to Datadog. Datadog’s OTLP metrics intake endpoint requires delta metrics, and host metadata sent to this endpoint does not populate the Datadog Infrastructure Host List. Before relying on direct OTel HTTP in production, verify that your expected Aerospike metrics, tags, monitors, and dashboards behave as expected. ::: note In some cases, Datadog may charge more for direct OTel HTTP than Datadog Agent. Contact Datadog support for further pricing information. ::: ## Connect and configure Datadog integration - [Datadog Agent](#tab-panel-2537) - [Direct OTel HTTP](#tab-panel-2538) ### Datadog Agent prerequisites You need Aerospike Database Enterprise Edition, Aerospike Prometheus Exporter, a Datadog account, and Datadog Agent installed on the host or container that can reach the exporter endpoint. The Datadog Agent must have a valid [Datadog API key](https://docs.datadoghq.com/account_management/api-app-keys/) and `site` value in `datadog.yaml` for your Datadog site. The Datadog account must have permission to install integrations. If your organization uses the `us2.ddog-gov.com` Datadog site, confirm that the **Aerospike Enterprise** integration is available in your organization or request enablement through Datadog support. For more information, see Datadog’s [Aerospike Enterprise integration documentation](https://docs.datadoghq.com/integrations/aerospike-enterprise/). If Datadog Agent and Aerospike Prometheus Exporter run on the same host, the default exporter endpoint is `http://127.0.0.1:9145/metrics`. If they run on different hosts, use a network-reachable hostname or IP address instead. ![Aerospike cluster metrics collected by Aerospike Prometheus Exporter and scraped by Datadog Agent before being sent to Datadog](https://aerospike.com/docs/_astro/aerospike-cluster-datadog-agent.D8BoUVT0_Z2u3WhC.png) ### Configure Datadog Agent 1. Install the Aerospike Enterprise Agent integration on each Datadog Agent that will collect Aerospike metrics. Terminal window ```bash datadog-agent integration install -t datadog-aerospike_enterprise==1.0.0 ``` 2. Create or edit the Datadog integration configuration file: ```text conf.d/aerospike_enterprise.d/conf.yaml ``` Set `openmetrics_endpoint` to the Aerospike Prometheus Exporter metrics endpoint. Replace _`APE_HOST`_ with the hostname or IP address reachable from the Datadog Agent. ```yaml init_config: instances: - openmetrics_endpoint: http://APE_HOST:9145/metrics ``` If the Datadog Agent and exporter run on the same host, use the loopback endpoint instead: ```yaml init_config: instances: - openmetrics_endpoint: http://127.0.0.1:9145/metrics ``` If your exporter endpoint uses TLS or authentication, add the matching OpenMetrics options from Datadog’s [sample `aerospike_enterprise.d/conf.yaml`](https://github.com/DataDog/integrations-extras/blob/master/aerospike_enterprise/datadog_checks/aerospike_enterprise/data/conf.yaml.example). 3. Restart the Datadog Agent after you save `conf.yaml`. 4. In Datadog, open the **Integrations** catalog and install **Aerospike Enterprise**. This installs the Datadog-side assets for the integration, including the integration tile and bundled dashboards. You can also query Aerospike metrics directly in Datadog by metric name, such as `aerospike.server.namespace.client_read_success.count` and `aerospike.server.latencies.read_ms_bucket`. After the metrics reach Datadog, you can explore them in the Datadog UI in **Metrics Explorer**, create monitors, and use the **Aerospike Enterprise** dashboards. ### Verify Datadog Agent 1. From the Datadog Agent host, confirm that the exporter endpoint returns Aerospike metrics. Command Terminal window ```bash curl http://APE_HOST:9145/metrics ``` Expected output ```text # HELP aerospike_... # TYPE aerospike_... aerospike_... ``` If this command cannot connect or does not return Aerospike metrics, Datadog cannot scrape the exporter. 2. Check the Datadog Agent status and look for `aerospike_enterprise` in the **Checks** section. Command Terminal window ```bash datadog-agent status ``` Expected output excerpt ```text Checks ====== aerospike_enterprise Instance ID: aerospike_enterprise:... ... ``` 3. In Datadog, open **Metrics Explorer** and search for metrics with the `aerospike.server.` prefix. Confirm that Datadog shows Aerospike metrics such as `aerospike.server.namespace.client_read_success.count` or `aerospike.server.latencies.read_ms_bucket`. Then open the **Aerospike Enterprise** dashboards and confirm that cluster and namespace widgets are receiving data. ### Troubleshoot Datadog Agent These steps focus on the Datadog configuration that affects Aerospike metric collection. For Datadog Agent installation, account, or dashboard issues that are not specific to Aerospike metrics, see Datadog’s troubleshooting documentation or support. #### The Agent cannot reach the exporter endpoint 1. From the Datadog Agent host, run `curl` against the same endpoint configured in `openmetrics_endpoint` in `conf.d/aerospike_enterprise.d/conf.yaml`. Command Terminal window ```bash curl http://APE_HOST:9145/metrics ``` Expected output ```text # HELP aerospike_... # TYPE aerospike_... aerospike_... ``` If the exporter endpoint uses TLS or basic authentication, use the matching options for that endpoint: Terminal window ```bash curl --cacert CA_FILE_PATH https://APE_HOST:9145/metrics curl -u DATADOG_AGENT_USERNAME:DATADOG_AGENT_PASSWORD http://APE_HOST:9145/metrics curl --cacert CA_FILE_PATH -u DATADOG_AGENT_USERNAME:DATADOG_AGENT_PASSWORD https://APE_HOST:9145/metrics ``` 2. If `curl` times out or returns a connection error, check that Aerospike Prometheus Exporter is running on the exporter host. Terminal window ```bash systemctl status aerospike-prometheus-exporter ``` 3. If the exporter is running, check that the Agent host can reach the exporter host on port `9145`. For example, run `curl` again with the hostname or IP address used in `openmetrics_endpoint`, not `127.0.0.1`. #### The Agent status does not show `aerospike_enterprise` 1. On the Datadog Agent host, run the integration install command again and check for install errors. Terminal window ```bash datadog-agent integration install -t datadog-aerospike_enterprise==1.0.0 ``` 2. Open the Agent configuration directory and check that the integration configuration file exists at `conf.d/aerospike_enterprise.d/conf.yaml`. 3. In `conf.d/aerospike_enterprise.d/conf.yaml`, check that `instances` contains an `openmetrics_endpoint` value. ```yaml instances: - openmetrics_endpoint: http://APE_HOST:9145/metrics ``` 4. Restart the Datadog Agent, then run `datadog-agent status` and look for `aerospike_enterprise` in the **Checks** section. #### The Agent check runs but metrics do not appear in Datadog 1. From the Datadog Agent host, run `curl` against the same URL configured for `openmetrics_endpoint` in `conf.d/aerospike_enterprise.d/conf.yaml`. The endpoint should return Aerospike metrics. 2. In the Datadog Agent main configuration file, `datadog.yaml`, check that `api_key` is set and that `site` matches your Datadog organization, such as `datadoghq.com`, `us3.datadoghq.com`, or `datadoghq.eu`. 3. From the Agent host, check outbound connectivity to your Datadog site. If your Agent uses a site other than `datadoghq.com`, replace `api.datadoghq.com` with the API hostname for that site. Command Terminal window ```bash curl -I https://api.datadoghq.com ``` Example output ```text HTTP/... ``` The HTTP status can vary because this command does not include an API request. A response from Datadog confirms network connectivity. A timeout, DNS failure, or connection refusal means the Agent host cannot reach Datadog. 4. Run `datadog-agent status` and check the `aerospike_enterprise` check output for collection errors. #### The exporter endpoint uses TLS or authentication 1. In `conf.d/aerospike_enterprise.d/conf.yaml`, update `openmetrics_endpoint` to use `https://` if the exporter serves metrics over TLS. ```yaml instances: - openmetrics_endpoint: https://APE_HOST:9145/metrics ``` 2. If the exporter uses a private certificate authority, add the certificate path with `tls_ca_cert` in `conf.d/aerospike_enterprise.d/conf.yaml`. ```yaml instances: - openmetrics_endpoint: https://APE_HOST:9145/metrics tls_ca_cert: CA_FILE_PATH ``` 3. If the exporter requires basic authentication, add `username` and `password` in `conf.d/aerospike_enterprise.d/conf.yaml`. ```yaml instances: - openmetrics_endpoint: https://APE_HOST:9145/metrics username: DATADOG_AGENT_USERNAME password: DATADOG_AGENT_PASSWORD ``` 4. Restart the Datadog Agent and check `datadog-agent status` for `aerospike_enterprise` collection errors. For additional OpenMetrics options, see Datadog’s [sample `aerospike_enterprise.d/conf.yaml`](https://github.com/DataDog/integrations-extras/blob/master/aerospike_enterprise/datadog_checks/aerospike_enterprise/data/conf.yaml.example). ### Direct OTel HTTP prerequisites You need Aerospike Database Enterprise Edition, Aerospike Prometheus Exporter 1.31.0 or later, a Datadog account, and a Datadog API key for the target Datadog site. The exporter must be able to make outbound HTTPS connections to the Datadog OTLP metrics intake endpoint. ::: note Protect the Datadog API key as a secret. Do not commit it to source control, use deployment secret management where available, restrict read access to `ape.toml` to the exporter runtime user, and rotate the key if it is exposed. For more information, see Datadog’s [API key documentation](https://docs.datadoghq.com/account_management/api-app-keys/). ::: The exporter sends metrics to Datadog directly, so Datadog Agent does not need to reach the exporter endpoint. ![Aerospike cluster metrics collected by Aerospike Prometheus Exporter and sent directly to Datadog through the OTLP metrics intake endpoint](https://aerospike.com/docs/_astro/aerospike-cluster-otel-mode.-WM0_HcR_ZxLD7P.png) ### Configure direct OTel HTTP 1. Create or edit the Aerospike Prometheus Exporter configuration file: ```text /etc/aerospike-prometheus-exporter/ape.toml ``` 2. In the `[Agent]` and `[Agent.OpenTelemetry]` sections, enable OTel and configure the Datadog OTLP metrics intake endpoint. Replace _`DATADOG_API_KEY`_ with your Datadog API key. The API key and `http_endpoint` must belong to the same Datadog site; for non-US1 sites, replace the endpoint with that site’s OTLP metrics endpoint from Datadog’s [OTLP metrics intake documentation](https://docs.datadoghq.com/opentelemetry/setup/otlp_ingest/metrics/). ```toml [Agent] enable_open_telemetry = true [Agent.OpenTelemetry] service_name = "aerospike-server-metrics-service" http_endpoint = "https://otlp.datadoghq.com/v1/metrics" headers = {"dd-api-key"="DATADOG_API_KEY"} counter_temporality = "delta" metric_name_prefix = "aerospike.server" metric_context_separator = "period" ``` The Datadog OTLP metrics intake endpoint accepts delta metrics. Aerospike Prometheus Exporter uses `delta` as the default counter temporality, but setting `counter_temporality` explicitly makes the Datadog requirement visible in the exporter configuration. 3. Restart Aerospike Prometheus Exporter after you save `ape.toml`. Use the restart command for your operating system or container platform. After the exporter sends metrics to Datadog, you can explore them in the Datadog UI in **Metrics Explorer** and create monitors. Verify the Aerospike Enterprise dashboards before relying on them with direct OTel HTTP because this path does not use the Datadog Agent integration check. ### Verify direct OTel HTTP 1. Check the Aerospike Prometheus Exporter logs after restart. Confirm that the exporter starts without OTel configuration errors and does not report Datadog export failures. If the exporter logs HTTP status errors from Datadog, verify the OTLP endpoint, API key, and outbound network access. 2. In Datadog, open **Metrics Explorer** and search for metrics with the `aerospike.server.` prefix. Confirm that Datadog shows Aerospike metrics such as `aerospike.server.namespace.client_read_success.count` or `aerospike.server.latencies.read_ms_bucket`. 3. Open the Datadog dashboards or monitors that you expect to use with Aerospike metrics. Confirm that the widgets receive data and that the tags you use for filtering and grouping are present. Host metadata from direct OTLP metrics does not populate the Datadog Infrastructure Host List. ### Troubleshoot direct OTel HTTP These steps focus on the Aerospike Prometheus Exporter settings and network checks required to send Aerospike metrics to Datadog. For Datadog account, API key management, or monitor behavior that is not specific to Aerospike metrics, see Datadog’s troubleshooting documentation or support. #### Metrics do not appear in Datadog 1. In Datadog, open **Organization Settings > API Keys** and check that the key used in `ape.toml` exists and is active. 2. In `ape.toml`, check that `http_endpoint` uses the Datadog site for your organization, such as `https://otlp.datadoghq.com/v1/metrics` for the US1 site. ```toml http_endpoint = "https://otlp.datadoghq.com/v1/metrics" ``` 3. If your organization uses a different Datadog site, replace `http_endpoint` in `ape.toml` with that site’s OTLP metrics intake endpoint. For example, use `https://otlp.datadoghq.eu/v1/metrics` for the EU1 site. 4. From the exporter host, check outbound HTTPS access to the OTLP metrics endpoint. If your organization uses a Datadog site other than US1, replace the URL with the value configured in `http_endpoint` in `ape.toml`. Command Terminal window ```bash curl -I https://otlp.datadoghq.com/v1/metrics ``` Expected output ```text HTTP/2 405 ``` The HTTP status can vary because this command does not send an OTLP metrics payload. A response from Datadog confirms network connectivity. A timeout, DNS failure, or connection refusal means the exporter host cannot reach the OTLP endpoint. 5. Check the exporter logs for startup errors, OTel configuration errors, and HTTP status responses from Datadog. #### Datadog rejects metrics or datapoints are missing 1. In `ape.toml`, check that the exporter sends delta metrics. ```toml counter_temporality = "delta" ``` Datadog rejects cumulative metrics sent to the OTLP metrics intake endpoint. After you change `counter_temporality`, restart Aerospike Prometheus Exporter and check the exporter logs for Datadog export errors. 2. In `ape.toml`, check that the metric name prefix matches your Datadog queries, monitors, and dashboards. ```toml metric_name_prefix = "aerospike.server" ``` 3. In Datadog **Metrics Explorer**, search for the configured prefix. If you change `metric_name_prefix`, update any dashboards and monitors that query the previous metric names. 4. Open a metric in **Metrics Explorer** and review the available tags. If a monitor or dashboard filters by a tag that is not present on the metric, update the filter or configure the exporter to send the tag in `ape.toml`. #### Datadog returns `403 Forbidden` 1. In Datadog, open **Organization Settings > API Keys** and check that the API key used in `ape.toml` is active. 2. Compare the Datadog site where you created the API key with the hostname in `http_endpoint` in `ape.toml`. For example, an API key created in the US1 site must use `https://otlp.datadoghq.com/v1/metrics`. 3. Check the exporter logs after restart. If `403 Forbidden` continues, create a new API key for the target Datadog site and replace the `dd-api-key` header value in `ape.toml`. #### Datadog returns `413 Request Entity Too Large` 1. Check whether the exporter logs show payloads that exceed the Datadog OTLP metrics intake limit. Datadog returns `413 Request Entity Too Large` when an OTLP request payload is larger than 500 KB uncompressed or 5 MB compressed after decompression. 2. Aerospike Prometheus Exporter does not expose an OTLP batch-size setting in `ape.toml`. To reduce the payload size, reduce the number of exported metrics with Aerospike Prometheus Exporter allowlist or blocklist settings. 3. In `ape.toml`, remove labels that you do not use in Datadog queries, monitors, or dashboards from the `[Agent]` `labels` setting. 4. Restart Aerospike Prometheus Exporter and check the exporter logs for another `413 Request Entity Too Large` response. #### Datadog shows missing or lower-than-expected datapoints 1. In **Metrics Explorer**, open the affected metric and compare the tags on datapoints that arrive at the same time. 2. If multiple datapoints for the same metric share the same timestamp and tag set, add distinguishing labels in the `[Agent]` `labels` setting in `ape.toml`, such as a cluster, environment, or exporter identifier. 3. Recheck the metric in **Metrics Explorer** after the next exporter push interval. Datadog keeps only the last datapoint for duplicate timestamp-plus-dimensions combinations. ## Next steps - [Monitoring stack components](https://aerospike.com/docs/database/observe/monitor/components) - [Configure Aerospike Prometheus Exporter](https://aerospike.com/docs/database/observe/monitor/configure/exporter) - [Configure Prometheus and AlertManager](https://aerospike.com/docs/database/observe/monitor/configure/prometheus) - [Datadog Aerospike Enterprise integration](https://docs.datadoghq.com/integrations/aerospike-enterprise/) - [Datadog Agent integration configuration](https://docs.datadoghq.com/getting_started/integrations/#configuring-agent-integrations) - [Datadog OTLP metrics intake endpoint](https://docs.datadoghq.com/opentelemetry/setup/otlp_ingest/metrics)