Introduce ringbuf queue in access log module (#170)

apache · Dec 16, 2024 · 023d6c2 · 023d6c2
1 parent 089d9c9
commit 023d6c2
Show file tree

Hide file tree

Showing 17 changed files with 324 additions and 107 deletions.
diff --git a/.github/workflows/rover.yaml b/.github/workflows/rover.yaml
@@ -342,7 +342,7 @@ jobs:
         if: ${{ failure() }}
         name: Upload Logs
         with:
-          name: logs
+          name: logs ${{ matrix.test.name }}
           path: "${{ env.SW_INFRA_E2E_LOG_DIR }}"
 
   required:

diff --git a/CHANGES.md b/CHANGES.md
@@ -16,6 +16,7 @@ Release Notes.
 * Support parallel parsing protocol data in the access log module.
 * Upgrade Go library to `1.22`, eBPF library to `0.16.0`.
 * Reduce missing details issue in the access log module.
+* Introduce ringbuf queue to improve performance in the access log module.
 
 #### Bug Fixes
 * Fix the base image cannot run in the arm64.

diff --git a/bpf/accesslog/common/connection.h b/bpf/accesslog/common/connection.h
@@ -21,6 +21,7 @@
 #include "socket.h"
 #include "data_args.h"
 #include "socket_opts.h"
+#include "queue.h"
 
 // syscall:connect
 struct connect_args_t {
@@ -106,19 +107,7 @@ struct socket_connect_event_t {
     __u64 conntrack_upstream_iph;
     __u32 conntrack_upstream_port;
 };
-struct {
-	__uint(type, BPF_MAP_TYPE_PERF_EVENT_ARRAY);
-} socket_connection_event_queue SEC(".maps");
-struct {
-    __uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
-    __type(key, __u32);
-    __type(value, struct socket_connect_event_t);
-    __uint(max_entries, 1);
-} socket_connect_event_per_cpu_map SEC(".maps");
-static __inline struct socket_connect_event_t* create_socket_connect_event() {
-  __u32 kZero = 0;
-  return bpf_map_lookup_elem(&socket_connect_event_per_cpu_map, &kZero);
-}
+DATA_QUEUE(socket_connection_event_queue, 1024 * 1024);
 
 // active connection cached into the hashmap
 // if connection closed, then deleted
@@ -170,9 +159,7 @@ struct socket_close_event_t {
     // close success
     __u32 success;
 };
-struct {
-	__uint(type, BPF_MAP_TYPE_PERF_EVENT_ARRAY);
-} socket_close_event_queue SEC(".maps");
+DATA_QUEUE(socket_close_event_queue, 1024 * 1024);
 
 static __inline bool family_should_trace(const __u32 family) {
     return family != AF_UNKNOWN && family != AF_INET && family != AF_INET6 ? false : true;
@@ -182,7 +169,8 @@ static __always_inline void submit_new_connection(void* ctx, bool success, __u32
                                             struct sockaddr* addr, const struct socket* socket, struct connect_track_remote* conntrack, __u8 role) {
     // send to the user-space the connection event
     __u64 curr_nacs = bpf_ktime_get_ns();
-    struct socket_connect_event_t *event = create_socket_connect_event();
+    struct socket_connect_event_t *event;
+    event = rover_reserve_buf(&socket_connection_event_queue, sizeof(*event));
     if (event == NULL) {
         return;
     }
@@ -195,12 +183,11 @@ static __always_inline void submit_new_connection(void* ctx, bool success, __u32
     event->end_time = curr_nacs;
     event->func_name = func_name;
     if (func_name == SOCKET_OPTS_TYPE_CONNECT) {
-        event->role = CONNECTION_ROLE_TYPE_CLIENT;
+        role = CONNECTION_ROLE_TYPE_CLIENT;
     } else if (func_name == SOCKET_OPTS_TYPE_ACCEPT) {
-        event->role = CONNECTION_ROLE_TYPE_SERVER;
-    } else {
-        event->role = role;
+        role = CONNECTION_ROLE_TYPE_SERVER;
     }
+    event->role = role;
     event->pid = tgid;
     event->sockfd = fd;
 
@@ -216,6 +203,7 @@ static __always_inline void submit_new_connection(void* ctx, bool success, __u32
     event->success = success;
 
     __u16 port;
+    __u8 socket_family;
     event->local_port = 0;
     event->remote_port = 0;
     if (socket == NULL) {
@@ -236,6 +224,7 @@ static __always_inline void submit_new_connection(void* ctx, bool success, __u32
         short unsigned int skc_family;
         BPF_CORE_READ_INTO(&skc_family, s, __sk_common.skc_family);
         event->socket_family = skc_family;
+        socket_family = skc_family;
 
         if (event->socket_family == AF_INET) {
             BPF_CORE_READ_INTO(&port, s, __sk_common.skc_num);
@@ -254,6 +243,7 @@ static __always_inline void submit_new_connection(void* ctx, bool success, __u32
         }
     } else if (addr != NULL) {
         event->socket_family = _(addr->sa_family);
+        socket_family = event->socket_family;
         if (event->socket_family == AF_INET) {
             struct sockaddr_in *daddr = (struct sockaddr_in *)addr;
             bpf_probe_read(&event->remote_addr_v4, sizeof(event->remote_addr_v4), &daddr->sin_addr.s_addr);
@@ -270,21 +260,22 @@ static __always_inline void submit_new_connection(void* ctx, bool success, __u32
         }
     } else {
         event->socket_family = AF_UNKNOWN;
+        socket_family = AF_UNKNOWN;
     }
 
-    bpf_perf_event_output(ctx, &socket_connection_event_queue, BPF_F_CURRENT_CPU, event, sizeof(*event));
+    rover_submit_buf(ctx, &socket_connection_event_queue, event, sizeof(*event));
     if (success == false) {
         return;
     }
 
     // if connect success, then add the activate connection into the kernel
     // active connection save
     struct active_connection_t con = {};
-    con.random_id = event->random_id;
+    con.random_id = random_id;
     con.pid = tgid;
     con.sockfd = fd;
-    con.role = event->role;
-    con.socket_family = event->socket_family;
+    con.role = role;
+    con.socket_family = socket_family;
     bpf_map_update_elem(&active_connection_map, &conid, &con, 0);
 }
 
@@ -312,17 +303,21 @@ static __inline void submit_connection_when_not_exists(void *ctx, __u64 id, stru
 }
 
 static __inline void notify_close_connection(void* ctx, __u64 conid, struct active_connection_t* con, __u64 start_time, __u64 end_time, int ret) {
-    struct socket_close_event_t close_event = {};
+    struct socket_close_event_t *close_event;
+    close_event = rover_reserve_buf(&socket_close_event_queue, sizeof(*close_event));
+    if (close_event == NULL) {
+        return;
+    }
 
-    close_event.conid = conid;
-    close_event.random_id = con->random_id;
-    close_event.start_time = start_time;
-    close_event.end_time = end_time;
-    close_event.pid = con->pid;
-    close_event.sockfd = con->sockfd;
-    close_event.success = ret > 0 ? true : false;
+    close_event->conid = conid;
+    close_event->random_id = con->random_id;
+    close_event->start_time = start_time;
+    close_event->end_time = end_time;
+    close_event->pid = con->pid;
+    close_event->sockfd = con->sockfd;
+    close_event->success = ret > 0 ? true : false;
 
-    bpf_perf_event_output(ctx, &socket_close_event_queue, BPF_F_CURRENT_CPU, &close_event, sizeof(close_event));
+    rover_submit_buf(ctx, &socket_close_event_queue, close_event, sizeof(*close_event));
 }
 
 static __inline void submit_close_connection(void* ctx, __u32 tgid, __u32 fd, __u64 start_nacs, int ret) {

diff --git a/bpf/accesslog/syscalls/transfer.h b/bpf/accesslog/syscalls/transfer.h
@@ -19,6 +19,7 @@
 #include "socket_opts.h"
 #include "socket_data.h"
 #include "socket_reader.h"
+#include "queue.h"
 #include "protocol_analyzer.h"
 #include "../common/connection.h"
 #include "../common/data_args.h"
@@ -68,15 +69,7 @@ struct socket_detail_t {
     __u8 ssl;
 };
 
-struct {
-    __uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
-    __type(key, __u32);
-    __type(value, struct socket_detail_t);
-    __uint(max_entries, 1);
-} socket_detail_event_per_cpu_map SEC(".maps");
-struct {
-	__uint(type, BPF_MAP_TYPE_PERF_EVENT_ARRAY);
-} socket_detail_data_queue SEC(".maps");
+DATA_QUEUE(socket_detail_queue, 1024 * 1024);
 
 static __always_inline void process_write_data(void *ctx, __u64 id, struct sock_data_args_t *args, ssize_t bytes_count,
                                         __u32 data_direction, const bool vecs, __u8 func_name, bool ssl) {
@@ -144,8 +137,8 @@ static __always_inline void process_write_data(void *ctx, __u64 id, struct sock_
     // 1. when the SSL connection sends SSL(unencrypted) message
     // 2. when the not SSL connection sends plain data
     if (conn->ssl == ssl) {
-        __u32 kZero = 0;
-        struct socket_detail_t *detail = bpf_map_lookup_elem(&socket_detail_event_per_cpu_map, &kZero);
+        struct socket_detail_t *detail;
+        detail = rover_reserve_buf(&socket_detail_queue, sizeof(*detail));
         if (detail != NULL) {
             detail->connection_id = conid;
             detail->random_id = conn->random_id;
@@ -177,7 +170,7 @@ static __always_inline void process_write_data(void *ctx, __u64 id, struct sock_
             detail->l2_enter_queue_count = args->l2_enter_queue_count;
             detail->l4_package_rcv_from_queue_time = args->total_package_receive_from_queue_time;
 
-            bpf_perf_event_output(ctx, &socket_detail_data_queue, BPF_F_CURRENT_CPU, detail, sizeof(*detail));
+            rover_submit_buf(ctx, &socket_detail_queue, detail, sizeof(*detail));
         }
     }
 

diff --git a/bpf/include/queue.h b/bpf/include/queue.h
@@ -0,0 +1,61 @@
+// Licensed to Apache Software Foundation (ASF) under one or more contributor
+// license agreements. See the NOTICE file distributed with
+// this work for additional information regarding copyright
+// ownership. Apache Software Foundation (ASF) licenses this file to you under
+// the Apache License, Version 2.0 (the "License"); you may
+// not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     http://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing,
+// software distributed under the License is distributed on an
+// "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+// KIND, either express or implied.  See the License for the
+// specific language governing permissions and limitations
+// under the License.
+
+#pragma once
+
+#include "api.h"
+
+#define DATA_QUEUE(name, size)               \
+	struct {                                    \
+		__uint(type, BPF_MAP_TYPE_RINGBUF); \
+		__uint(max_entries, size);          \
+	} name SEC(".maps");                        \
+	const void *rover_data_queue_##name __attribute__((unused));
+
+struct {
+	__uint(type, BPF_MAP_TYPE_PERCPU_ARRAY);
+	__uint(max_entries, 1);
+	__uint(key_size, sizeof(__u32));
+	__uint(value_size, 10240); // all events are less than 10KB
+} rover_data_heap SEC(".maps");
+
+static __always_inline void *rover_reserve_buf(void *map, __u64 size) {
+	static const int zero = 0;
+
+	if (bpf_core_enum_value_exists(enum bpf_func_id,
+				       BPF_FUNC_ringbuf_reserve))
+		return bpf_ringbuf_reserve(map, size, 0);
+
+	return bpf_map_lookup_elem(&rover_data_heap, &zero);
+}
+
+static __always_inline void rover_discard_buf(void *buf)
+{
+	if (bpf_core_enum_value_exists(enum bpf_func_id,
+				       BPF_FUNC_ringbuf_discard))
+		bpf_ringbuf_discard(buf, 0);
+}
+
+static __always_inline long rover_submit_buf(void *ctx, void *map, void *buf, __u64 size) {
+	if (bpf_core_enum_value_exists(enum bpf_func_id,
+				       BPF_FUNC_ringbuf_submit)) {
+		bpf_ringbuf_submit(buf, 0);
+		return 0;
+	}
+
+	return bpf_perf_event_output(ctx, map, BPF_F_CURRENT_CPU, buf, size);
+}