This document was translated by ChatGPT

1. Deployment

1. What is the difference between all-in-one deployment mode and regular deployment?

   A: All-in-one means that the storage components clickhouse and mysql do not have corresponding PVCs and are deployed using the hostPath mode. If the K8S cluster has multiple nodes, after restarting the deepflow-clickhouse/mysql Pod, it may drift to another node, causing previously collected data to be unavailable for query. It is recommended to use all-in-one deployment for trial purposes, and use regular deployment mode for testing/POC phases.

2. How long is the data generally retained, and can it be adjusted?

   A: The retention period varies for different types of data. You can check the retention periods for different data types in server.yaml (opens new window), and adjust them before the first deployment by modifying the default configuration during helm installation.

3. How to use external MySQL/ClickHouse?

   A: See the sections Using Managed MySQL and Using Managed ClickHouse in Production Deployment Recommendations.

4. The deployment specification includes two storage components, mysql and clickhouse. What is the difference between them?

   A: mysql stores metadata obtained from the deployed cluster, such as virtual machines, K8S resources, and synchronized collector information. clickhouse stores real-time collected data, such as network flow logs collected from the cluster, and performs aggregation and analysis.

5. No data on Grafana after deployment?

   A: Please troubleshoot as follows:

   • Check whether all Pods are running normally: run kubectl get pods -n deepflow and confirm that all Pods are in the Running state.

   • Check whether DeepFlow Agent and DeepFlow Server are connected successfully: run deepflow-ctl domain list to check if a service domain has been successfully created, and run deepflow-ctl agent list to check whether the STATE is NORMAL.

   • If dashboards of the Network - X type have no data, check whether the NIC name matches the capture rules. You can run deepflow-ctl agent-group-config example | grep tap_interface_regex to view the default capture range. If you are using a custom CNI or have set up the network in another way, add the NIC matching rule to tap_interface_regex and update the agent configuration.

   • If only dashboards of the Application - X type have no data, confirm that the application protocols used in the cluster meet the supported list.

6. I have configured OpenTelemetry data integration / want to use DeepFlow's eBPF tracing and network tracing capabilities, but there is no data in the Distributed Tracing dashboard?

   A: Please troubleshoot as follows:

   • Using OpenTelemetry integration:

     • Confirm that the application has integrated the OTel SDK or started the OTel Agent.

     • Confirm that you have completed the configuration according to Configure DeepFlow. On the container node where deepflow-agent is located, run netstat -alntp | grep 38086 to check whether this feature has started normally. If configured, check in Network - Flow Log whether there are flow logs with Server Port 38086.

     • In the Application - K8s Pod Map dashboard, check whether there is traffic from the application to the otel-agent to the container node, ensuring that the network path is smooth and requests are occurring.

     • In the Application - Request Log dashboard, confirm whether there are any anomalies in the requests sent.

   • Using eBPF capabilities:

     • Confirm that the server kernel version meets the requirements.

     • Check all replicas of deepflow-agent: run kubectl logs -n deepflow ds/deepflow-agent | grep 'ebpf collector' to check whether the eBPF module has started normally, and run kubectl logs -n deepflow ds/deepflow-agent | grep TRACER to confirm that the eBPF Tracer function is running normally.

2. Product

1. After installation and deployment, what should I do next? Are there any product cases or usage scenarios to share?

   A: You can find our shared cases in the Getting Started with Observability (opens new window) blog series and the troubleshooting (opens new window) blog series. You can also review past sharing sessions on our Bilibili account (opens new window).

2. I think some features are not good enough and want to give suggestions. How can I do that?

   A: You are welcome to submit a Feature Request in Github Issue (opens new window). If you already have a mature idea, you can also put it into practice directly and submit it in Github PR (opens new window).

3. Where can I follow the latest updates of DeepFlow?

   A: You can check our latest release overview in the Release Notes or follow our latest blog (opens new window).

3. Contact Us

If the above information does not solve your problem, you can submit an issue via Github Issue (opens new window) or contact us directly (opens new window) for communication.