v6.1.8.6

Release v6.1.8.6

v6.2.2

New Features (GA)

• AutoTracing
  • Supports distributed tracing of Golang applications with zero interpolation
• Auto Tagging
  • Support adding custom metadata for processes, cloud servers, and K8s Namespace FR-001-Xiaomi
  • Support automatic synchronization of K8s cluster information under AWS and Alibaba Cloud accounts
• Management
  • Support the number of ClickHouse nodes greater than the number of deepflow-server replicas FR-003-Zhongtong
  • Support deepflow-agent running on K8s Node as a normal process (not Pod) FR-004-Tencent
  • Support specifying domain name controller or ingester address for deepflow-agent FR-008-Xiaomi

optimization

• deepflow-agent
  • The list of regular expressions for scanning processes (os-proc-regex) supports configuring action=drop to express ignore semantics FR-010-Xiaomi
  • Support running in Linux Kernel environment lower than 3.0 FR-012-Xiaomi
  • Use the socket information of the operating system to correct the flow log direction FR-011-Xiaomi
  • When the ctrl_ip or ctrl_mac of the agent's operating environment changes, it supports automatic update of the corresponding information of the agent
• deepflow-server
  • At the end of the UDP flow timeout, the status field of l4_flow_log is set to OK

v6.2.1

New Features (Alpha)

-Universal Application Topology

• Support zero interpolation code to automatically display the panoramic application topology of process granularity FR-001-XIAOMI
  -Auto Tagging
• Support adding custom metadata for processes, cloud servers, and K8s Namespace FR-001-XIAOMI
• Support automatic synchronization of K8s cluster information under AWS and Alibaba Cloud accounts
• Supports synchronization of Baidu Cloud Smart Network (CSN) resource information
• Querier API
  • Support PromQL
• Management
  • Support the number of ClickHouse nodes greater than the number of deepflow-server replicas FR-003-ZHONGTONG
  • Support deepflow-agent running on K8s Node as a normal process (not Pod) FR-004-Tencent
  • Support specifying domain name controller or ingester address for deepflow-agent FR-008-XIAOMI

optimization

• Querier API
  • Support returning the original field name before AS
• Grafana
  • Optimize the Variable of Enum type to avoid expanding all candidate values in SQL when selecting All

v6.1.8.5

Release v6.1.8.5

v6.1.8.4

Release v6.1.8.4

v6.2.0

Release v6.2.0

v6.1.8.3

Release v6.1.8.3

v6.1.8.2

new features

• AutoMetrics, AutoTracing, AutoLogging
  • Added support for SOFARPC protocol
• Grafana
  • Support as a Grafana Tempo data source and display Tracing data on the Tempo page

v6.1.8.1

new features

• AutoMetrics
  • Pre-aggregate eBPF Request Log to supplement tap_side = [client process/server process](https://deepflow.yunshan.net/docs/zh/auto-metrics/metrics-without-instrumentation/#statistical position Description) app performance metrics data for

v6.1.8

new features

• AutoMetrics, AutoTracing, AutoLogging
  • Added support for the ProtobufRPC protocol cluster, the current version supports the KRPC protocol
• Management
  • Add a signal_source tag field to the data table to indicate the signal source. Currently, the three signal sources supported for distinction are: Packet (cBPF), eBPF, OTel
  • deepflow-agent supports running in K8s macvlan/ipvlan CNI environment, and automatically calculates Pod access relationship and performance indicators, detailed configuration method of macvlan environment [reference document](https://deepflow.yunshan.net/docs/zh/ install/advanced-config/agent-advanced-config/#macvlan)
  • Support deepflow-agent to run in K8s Pod without SYS_ADMIN permission, detailed permission requirements [reference document](https://deepflow.yunshan.net/docs/zh/install/overview/#Run permission and kernel requirements )
  • Support configuring deepflow-agent to collect the [maximum length] of eBPF Socket Data (https://github.com/deepflowys/deepflow/blob/main/server/controller/model/agent_group_config_example.yaml#L252) to improve collection performance
  • Support configuration of [resource information] (https://github.com/deepflowys/deepflow/blob/main/server/server.yaml#L28) synchronized to the ingester module in deepflow-server, thereby reducing the ingester module under multiple K8s clusters Resource consumption of the AutoTagging feature

optimization

• Management
  • Reduce the memory overhead of deepflow-agent when watching K8s apiserver through message compression and clipping
  • Optimize the memory consumption of deepflow-server when calculating service Tag through Pod aggregation
  • Optimize the write pressure of deepflow-server on the flow_tag database by means of memory pre-aggregation
  • Managed MySQL service supports connections using non-root users