Adding upstream version 1.44.3.upstream/1.44.3upstream

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

1 files changed, 692 insertions, 0 deletions

diff --git a/fluent-bit/src/flb_chunk_trace.c b/fluent-bit/src/flb_chunk_trace.c
new file mode 100644
index 00000000..adacb73f
--- /dev/null
+++ b/fluent-bit/src/flb_chunk_trace.c
@@ -0,0 +1,692 @@
+/* -*- Mode: C; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
+
+/*  Fluent Bit
+ *  ==========
+ *  Copyright (C) 2015-2022 The Fluent Bit Authors
+ *
+ *  Licensed 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.
+ */
+
+#include <fcntl.h>
+
+#include <msgpack.h>
+#include <chunkio/chunkio.h>
+
+#include <fluent-bit/flb_info.h>
+#include <fluent-bit/flb_input_chunk.h>
+#include <fluent-bit/flb_input.h>
+#include <fluent-bit/flb_output.h>
+#include <fluent-bit/flb_chunk_trace.h>
+#include <fluent-bit/flb_pack.h>
+#include <fluent-bit/flb_base64.h>
+#include <fluent-bit/flb_storage.h>
+#include <fluent-bit/flb_router.h>
+#include <fluent-bit/flb_kv.h>
+
+
+/* Register external function to emit records, check 'plugins/in_emitter' */
+int in_emitter_add_record(const char *tag, int tag_len,
+                          const char *buf_data, size_t buf_size,
+                          struct flb_input_instance *in);
+
+/****************************************************************************/
+/* To avoid double frees when enabling and disabling tracing as well        */
+/* as avoiding race conditions when stopping fluent-bit while someone is    */
+/* toggling tracing via the HTTP API this set of APIS with a mutex lock     */
+/* is used:                                                                 */
+/*   * flb_chunk_trace_to_be_destroyed - query to see if the trace context  */
+/*     is slated to be freed                                                */
+/*   * flb_chunk_trace_set_destroy - set the trace context to be destroyed  */
+/*     once all chunks are freed (executed in flb_chunk_trace_destroy).     */
+/*   * flb_chunk_trace_has_chunks - see if there are still chunks using     */
+/*     using the tracing context                                            */
+/*   * flb_chunk_trace_add - increment the traces chunk count               */
+/*   * flb_chunk_trace_sub - decrement the traces chunk count               */
+/****************************************************************************/
+static inline int flb_chunk_trace_to_be_destroyed(struct flb_chunk_trace_context *ctxt)
+{
+    int ret = FLB_FALSE;
+
+    ret = (ctxt->to_destroy == 1 ? FLB_TRUE : FLB_FALSE);
+    return ret;
+}
+
+static inline int flb_chunk_trace_has_chunks(struct flb_chunk_trace_context *ctxt)
+{
+    int ret = FLB_FALSE;
+
+    ret = ((ctxt->chunks > 0) ? FLB_TRUE : FLB_FALSE);
+    return ret;
+}
+
+static inline void flb_chunk_trace_add(struct flb_chunk_trace_context *ctxt)
+{
+    ctxt->chunks++;
+}
+
+static inline void flb_chunk_trace_sub(struct flb_chunk_trace_context *ctxt)
+{
+    ctxt->chunks--;
+}
+
+static inline void flb_chunk_trace_set_destroy(struct flb_chunk_trace_context *ctxt)
+{
+    ctxt->to_destroy = 1;
+}
+
+static struct flb_output_instance *find_calyptia_output_instance(struct flb_config *config)
+{
+    struct mk_list *head = NULL;
+    struct flb_output_instance *output = NULL;
+
+    mk_list_foreach(head, &config->outputs) {
+        output = mk_list_entry(head, struct flb_output_instance, _head);
+        if (strcmp(output->p->name, "calyptia") == 0) {
+            return output;
+        }
+    }
+    return NULL;
+}
+
+static void trace_chunk_context_destroy(struct flb_chunk_trace_context *ctxt)
+{
+    int i;
+
+
+    if (flb_chunk_trace_has_chunks(ctxt) == FLB_TRUE) {
+        flb_chunk_trace_set_destroy(ctxt);
+        flb_input_pause_all(ctxt->flb->config);
+        return;
+    }
+    
+    /* pause all inputs, then destroy the input storage. */
+    flb_input_pause_all(ctxt->flb->config);
+    /* waiting for all tasks to end is key to safely stopping and destroying */
+    /* the fluent-bit pipeline. */
+    for (i = 0; i < 5 && flb_task_running_count(ctxt->flb->config) > 0; i++) {
+        usleep(10 * 1000);
+    }
+
+    flb_sds_destroy(ctxt->trace_prefix);
+    flb_stop(ctxt->flb);
+    flb_destroy(ctxt->flb);
+    flb_free(ctxt);
+}
+
+void flb_chunk_trace_context_destroy(void *input)
+{
+    struct flb_input_instance *in = (struct flb_input_instance *)input;
+    pthread_mutex_lock(&in->chunk_trace_lock);
+    if (in->chunk_trace_ctxt != NULL) {
+        trace_chunk_context_destroy(in->chunk_trace_ctxt);
+        in->chunk_trace_ctxt = NULL;
+    }
+    pthread_mutex_unlock(&in->chunk_trace_lock);
+}
+
+struct flb_chunk_trace_context *flb_chunk_trace_context_new(void *trace_input,
+                                                            const char *output_name,
+                                                            const char *trace_prefix,
+                                                            void *data, struct mk_list *props)
+{
+    struct flb_input_instance *in = (struct flb_input_instance *)trace_input;
+    struct flb_config *config = in->config;
+    struct flb_input_instance *input = NULL;
+    struct flb_output_instance *output = NULL;
+    struct flb_output_instance *calyptia = NULL;
+    struct flb_chunk_trace_context *ctx = NULL;
+    struct mk_list *head = NULL;
+    struct flb_kv *prop = NULL;
+    int ret;
+
+    if (config->enable_chunk_trace == FLB_FALSE) {
+        flb_warn("[chunk trace] enable chunk tracing via the configuration or "
+                 " command line to be able to activate tracing.");
+        return NULL;
+    }
+
+    pthread_mutex_lock(&in->chunk_trace_lock);
+
+    if (in->chunk_trace_ctxt) {
+        trace_chunk_context_destroy(in->chunk_trace_ctxt);
+    }
+
+    ctx = flb_calloc(1, sizeof(struct flb_chunk_trace_context));
+    if (ctx == NULL) {
+        flb_errno();
+        pthread_mutex_unlock(&in->chunk_trace_lock);
+        return NULL;
+    }
+
+    ctx->flb = flb_create();
+    if (ctx->flb == NULL) {
+        flb_errno();
+        goto error_ctxt;
+    }
+
+    flb_service_set(ctx->flb, "flush", "1", "grace", "1", NULL);
+
+    input = (void *)flb_input_new(ctx->flb->config, "emitter", NULL, FLB_FALSE);
+    if (input == NULL) {
+        flb_error("could not load trace emitter");
+        goto error_flb;
+    }
+
+    ret = flb_input_set_property(input, "alias", "trace-emitter");
+    if (ret != 0) {
+        flb_error("unable to set alias for trace emitter");
+        goto error_input;
+    }
+
+    ret = flb_input_set_property(input, "ring_buffer_size", "4096");
+    if (ret != 0) {
+        flb_error("unable to set ring buffer size for trace emitter");
+        goto error_input;
+    }
+
+    output = flb_output_new(ctx->flb->config, output_name, data, 1);
+    if (output == NULL) {
+        flb_error("could not create trace output");
+        goto error_input;
+    }
+    
+    /* special handling for the calyptia plugin so we can copy the API */
+    /* key and other configuration properties. */
+    if (strcmp(output_name, "calyptia") == 0) {
+        calyptia = find_calyptia_output_instance(config);
+        if (calyptia == NULL) {
+            flb_error("unable to find calyptia output instance");
+            goto error_output;
+        }
+        mk_list_foreach(head, &calyptia->properties) {
+            prop = mk_list_entry(head, struct flb_kv, _head);
+            flb_output_set_property(output, prop->key, prop->val);
+        }        
+    }
+    else if (props != NULL) {
+        mk_list_foreach(head, props) {
+            prop = mk_list_entry(head, struct flb_kv, _head);
+            flb_output_set_property(output, prop->key, prop->val);
+        }
+    }
+
+    ret = flb_router_connect_direct(input, output);
+    if (ret != 0) {
+        flb_error("unable to route traces");
+        goto error_output;
+    }
+
+    ctx->output = (void *)output;
+    ctx->input = (void *)input;
+    ctx->trace_prefix = flb_sds_create(trace_prefix);
+
+    flb_start_trace(ctx->flb);
+
+    in->chunk_trace_ctxt = ctx;
+    pthread_mutex_unlock(&in->chunk_trace_lock);
+    return ctx;
+
+error_output:
+    flb_output_instance_destroy(output);
+error_input:
+    if (ctx->cio) {
+        cio_destroy(ctx->cio);
+    }
+    flb_input_instance_destroy(input);
+error_flb:
+    flb_destroy(ctx->flb);
+error_ctxt:
+    flb_free(ctx);
+    pthread_mutex_unlock(&in->chunk_trace_lock);
+    return NULL;
+}
+
+struct flb_chunk_trace *flb_chunk_trace_new(struct flb_input_chunk *chunk)
+{
+    struct flb_chunk_trace *trace = NULL;
+    struct flb_input_instance *f_ins = (struct flb_input_instance *)chunk->in;
+
+    pthread_mutex_lock(&f_ins->chunk_trace_lock);
+
+    if (flb_chunk_trace_to_be_destroyed(f_ins->chunk_trace_ctxt) == FLB_TRUE) {
+        pthread_mutex_unlock(&f_ins->chunk_trace_lock);
+        return NULL;
+    }
+
+    trace = flb_calloc(1, sizeof(struct flb_chunk_trace));
+    if (trace == NULL) {
+        flb_errno();
+        pthread_mutex_unlock(&f_ins->chunk_trace_lock);
+        return NULL;
+    }
+
+    trace->ctxt = f_ins->chunk_trace_ctxt;
+    flb_chunk_trace_add(trace->ctxt);
+
+    trace->trace_id = flb_sds_create("");
+    if (flb_sds_printf(&trace->trace_id, "%s%d", trace->ctxt->trace_prefix,
+                  trace->ctxt->trace_count++) == NULL) {
+        pthread_mutex_unlock(&f_ins->chunk_trace_lock);
+        flb_sds_destroy(trace->trace_id);
+        flb_free(trace);
+        return NULL;
+    }
+
+    trace->ic = chunk;
+
+    pthread_mutex_unlock(&f_ins->chunk_trace_lock);
+    return trace;
+}
+
+void flb_chunk_trace_destroy(struct flb_chunk_trace *trace)
+{
+    pthread_mutex_lock(&trace->ic->in->chunk_trace_lock);
+    flb_chunk_trace_sub(trace->ctxt);
+
+    /* check to see if we need to free the trace context. */
+    if (flb_chunk_trace_has_chunks(trace->ctxt) == FLB_FALSE &&
+        flb_chunk_trace_to_be_destroyed(trace->ctxt) == FLB_TRUE) {
+        trace_chunk_context_destroy(trace->ctxt);
+    }
+    else if (flb_chunk_trace_has_chunks(trace->ctxt) == FLB_TRUE &&
+        flb_chunk_trace_to_be_destroyed(trace->ctxt) == FLB_TRUE) {
+    }
+    pthread_mutex_unlock(&trace->ic->in->chunk_trace_lock);
+
+    flb_sds_destroy(trace->trace_id);
+    flb_free(trace);
+}
+
+int flb_chunk_trace_context_set_limit(void *input, int limit_type, int limit_arg)
+{
+    struct flb_input_instance *in = (struct flb_input_instance *)input;
+    struct flb_chunk_trace_context *ctxt = NULL;
+    struct flb_time tm;
+
+    pthread_mutex_lock(&in->chunk_trace_lock);
+
+    ctxt = in->chunk_trace_ctxt;
+    if (ctxt == NULL) {
+        pthread_mutex_unlock(&in->chunk_trace_lock);
+        return -1;
+    }
+
+    switch(limit_type) {
+    case FLB_CHUNK_TRACE_LIMIT_TIME:
+        flb_time_get(&tm);
+        ctxt->limit.type = FLB_CHUNK_TRACE_LIMIT_TIME;
+        ctxt->limit.seconds_started = tm.tm.tv_sec;
+        ctxt->limit.seconds = limit_arg;
+        
+        pthread_mutex_unlock(&in->chunk_trace_lock);
+        return 0;
+    case FLB_CHUNK_TRACE_LIMIT_COUNT:
+        ctxt->limit.type = FLB_CHUNK_TRACE_LIMIT_COUNT;
+        ctxt->limit.count = limit_arg;
+
+        pthread_mutex_unlock(&in->chunk_trace_lock);
+        return 0;
+    }
+
+    pthread_mutex_unlock(&in->chunk_trace_lock);
+    return -1;
+}
+
+int flb_chunk_trace_context_hit_limit(void *input)
+{
+    struct flb_input_instance *in = (struct flb_input_instance *)input;
+    struct flb_time tm;
+    struct flb_chunk_trace_context *ctxt = NULL;
+
+    pthread_mutex_lock(&in->chunk_trace_lock);
+
+    ctxt = in->chunk_trace_ctxt;
+    if (ctxt == NULL) {
+        pthread_mutex_unlock(&in->chunk_trace_lock);
+        return FLB_FALSE;
+    }
+
+    switch(ctxt->limit.type) {
+    case FLB_CHUNK_TRACE_LIMIT_TIME:
+        flb_time_get(&tm);
+        if ((tm.tm.tv_sec - ctxt->limit.seconds_started) > ctxt->limit.seconds) {
+            pthread_mutex_unlock(&in->chunk_trace_lock);
+            return FLB_TRUE;
+        }
+        return FLB_FALSE;
+    case FLB_CHUNK_TRACE_LIMIT_COUNT:
+        if (ctxt->limit.count <= ctxt->trace_count) {
+            pthread_mutex_unlock(&in->chunk_trace_lock);
+            return FLB_TRUE;
+        }
+        pthread_mutex_unlock(&in->chunk_trace_lock);
+        return FLB_FALSE;
+    }
+    pthread_mutex_unlock(&in->chunk_trace_lock);
+    return FLB_FALSE;
+}
+
+void flb_chunk_trace_do_input(struct flb_input_chunk *ic)
+{
+    pthread_mutex_lock(&ic->in->chunk_trace_lock);
+    if (ic->in->chunk_trace_ctxt == NULL) {
+        pthread_mutex_unlock(&ic->in->chunk_trace_lock);
+    	return;
+    }
+    pthread_mutex_unlock(&ic->in->chunk_trace_lock);
+    
+    if (ic->trace == NULL) {
+        ic->trace = flb_chunk_trace_new(ic);
+    }
+
+    if (ic->trace) {
+        flb_chunk_trace_input(ic->trace);
+        if (flb_chunk_trace_context_hit_limit(ic->in) == FLB_TRUE) {
+            flb_chunk_trace_context_destroy(ic->in);
+        }
+    }
+}
+
+int flb_chunk_trace_input(struct flb_chunk_trace *trace)
+{
+    msgpack_packer mp_pck;
+    msgpack_sbuffer mp_sbuf;
+    msgpack_unpacked result;
+    msgpack_object *record = NULL;
+    char *buf = NULL;
+    size_t buf_size;
+    struct flb_time tm;
+    struct flb_time tm_end;
+    struct flb_input_instance *input = (struct flb_input_instance *)trace->ic->in;
+    int rc = -1;
+    size_t off = 0;
+    flb_sds_t tag = flb_sds_create("trace");
+    int records = 0;
+
+
+    /* initiailize start time */
+    flb_time_get(&tm);
+    flb_time_get(&tm_end);
+
+    msgpack_sbuffer_init(&mp_sbuf);
+    msgpack_packer_init(&mp_pck, &mp_sbuf, msgpack_sbuffer_write);
+    msgpack_unpacked_init(&result);
+
+    cio_chunk_get_content(trace->ic->chunk, &buf, &buf_size);
+    
+    msgpack_pack_array(&mp_pck, 2);
+    flb_pack_time_now(&mp_pck);
+    if (input->alias != NULL) {
+        msgpack_pack_map(&mp_pck, 7);
+    } 
+    else {
+        msgpack_pack_map(&mp_pck, 6);
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "type", 4);
+    msgpack_pack_int(&mp_pck, FLB_CHUNK_TRACE_TYPE_INPUT);
+
+    msgpack_pack_str_with_body(&mp_pck, "trace_id", strlen("trace_id"));
+    msgpack_pack_str_with_body(&mp_pck, trace->trace_id, strlen(trace->trace_id));
+
+    msgpack_pack_str_with_body(&mp_pck, "plugin_instance", strlen("plugin_instance"));
+    msgpack_pack_str_with_body(&mp_pck, input->name, strlen(input->name));
+
+    if (input->alias != NULL) {
+        msgpack_pack_str_with_body(&mp_pck, "plugin_alias", strlen("plugin_alias"));
+        msgpack_pack_str_with_body(&mp_pck, input->alias, strlen(input->alias));
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "records", strlen("records"));
+
+    if (buf_size > 0) {
+        do {
+            rc = msgpack_unpack_next(&result, buf, buf_size, &off);
+            if (rc != MSGPACK_UNPACK_SUCCESS) {
+                flb_error("unable to unpack record");
+                goto sbuffer_error;
+            }
+            records++;
+        } while (rc == MSGPACK_UNPACK_SUCCESS && off < buf_size);
+
+        msgpack_pack_array(&mp_pck, records);
+
+        off = 0;
+        do {
+            rc = msgpack_unpack_next(&result, buf, buf_size, &off);
+            if (rc != MSGPACK_UNPACK_SUCCESS) {
+                flb_error("unable to unpack record");
+                goto sbuffer_error;
+            }
+            flb_time_pop_from_msgpack(&tm, &result, &record);
+
+            msgpack_pack_map(&mp_pck, 2);
+            msgpack_pack_str_with_body(&mp_pck, "timestamp", strlen("timestamp"));
+            flb_time_append_to_msgpack(&tm, &mp_pck, FLB_TIME_ETFMT_INT);
+            msgpack_pack_str_with_body(&mp_pck, "record", strlen("record"));
+            msgpack_pack_object(&mp_pck, *record);
+
+        } while (rc == MSGPACK_UNPACK_SUCCESS && off < buf_size);
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "start_time", strlen("start_time"));
+    flb_time_append_to_msgpack(&tm, &mp_pck, FLB_TIME_ETFMT_INT);
+    msgpack_pack_str_with_body(&mp_pck, "end_time", strlen("end_time"));
+    flb_time_append_to_msgpack(&tm_end, &mp_pck, FLB_TIME_ETFMT_INT);
+    in_emitter_add_record(tag, flb_sds_len(tag), mp_sbuf.data, mp_sbuf.size,
+                          trace->ctxt->input);
+sbuffer_error:
+    flb_sds_destroy(tag);
+    msgpack_unpacked_destroy(&result);
+    msgpack_sbuffer_destroy(&mp_sbuf);
+    return rc;
+}
+
+int flb_chunk_trace_pre_output(struct flb_chunk_trace *trace)
+{
+    msgpack_packer mp_pck;
+    msgpack_sbuffer mp_sbuf;
+    msgpack_unpacked result;
+    msgpack_object *record = NULL;
+    char *buf = NULL;
+    size_t buf_size;
+    struct flb_time tm;
+    struct flb_time tm_end;
+    struct flb_input_instance *input = (struct flb_input_instance *)trace->ic->in;
+    int rc = -1;
+    size_t off = 0;
+    flb_sds_t tag = flb_sds_create("trace");
+    int records = 0;
+
+
+    /* initiailize start time */
+    flb_time_get(&tm);
+    flb_time_get(&tm_end);
+
+    msgpack_sbuffer_init(&mp_sbuf);
+    msgpack_packer_init(&mp_pck, &mp_sbuf, msgpack_sbuffer_write);
+    msgpack_unpacked_init(&result);
+
+    cio_chunk_get_content(trace->ic->chunk, &buf, &buf_size);
+
+    msgpack_pack_array(&mp_pck, 2);
+    flb_pack_time_now(&mp_pck);
+    if (input->alias != NULL) {
+        msgpack_pack_map(&mp_pck, 7);
+    } 
+    else {
+        msgpack_pack_map(&mp_pck, 6);
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "type", 4);
+    msgpack_pack_int(&mp_pck, FLB_CHUNK_TRACE_TYPE_PRE_OUTPUT);
+
+    msgpack_pack_str_with_body(&mp_pck, "trace_id", strlen("trace_id"));
+    msgpack_pack_str_with_body(&mp_pck, trace->trace_id, strlen(trace->trace_id));
+
+    msgpack_pack_str_with_body(&mp_pck, "plugin_instance", strlen("plugin_instance"));
+    msgpack_pack_str_with_body(&mp_pck, input->name, strlen(input->name));
+
+    if (input->alias != NULL) {
+        msgpack_pack_str_with_body(&mp_pck, "plugin_alias", strlen("plugin_alias"));
+        msgpack_pack_str_with_body(&mp_pck, input->alias, strlen(input->alias));
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "records", strlen("records"));
+
+    if (buf_size > 0) {
+        do {
+            rc = msgpack_unpack_next(&result, buf, buf_size, &off);
+            if (rc != MSGPACK_UNPACK_SUCCESS) {
+                flb_error("unable to unpack record");
+                goto sbuffer_error;
+            }
+            records++;
+        } while (rc == MSGPACK_UNPACK_SUCCESS && off < buf_size);
+
+        msgpack_pack_array(&mp_pck, records);
+        off = 0;
+        do {
+            rc = msgpack_unpack_next(&result, buf, buf_size, &off);
+            if (rc != MSGPACK_UNPACK_SUCCESS) {
+                flb_error("unable to unpack record");
+                goto sbuffer_error;
+            }
+            flb_time_pop_from_msgpack(&tm, &result, &record);
+
+            msgpack_pack_map(&mp_pck, 2);
+            msgpack_pack_str_with_body(&mp_pck, "timestamp", strlen("timestamp"));
+            flb_time_append_to_msgpack(&tm, &mp_pck, FLB_TIME_ETFMT_INT);
+            msgpack_pack_str_with_body(&mp_pck, "record", strlen("record"));
+            msgpack_pack_object(&mp_pck, *record);
+
+        } while (rc == MSGPACK_UNPACK_SUCCESS && off < buf_size);
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "start_time", strlen("start_time"));
+    flb_time_append_to_msgpack(&tm, &mp_pck, FLB_TIME_ETFMT_INT);
+    msgpack_pack_str_with_body(&mp_pck, "end_time", strlen("end_time"));
+    flb_time_append_to_msgpack(&tm_end, &mp_pck, FLB_TIME_ETFMT_INT);
+    in_emitter_add_record(tag, flb_sds_len(tag), mp_sbuf.data, mp_sbuf.size,
+                          trace->ctxt->input);
+sbuffer_error:
+    flb_sds_destroy(tag);
+    msgpack_unpacked_destroy(&result);
+    msgpack_sbuffer_destroy(&mp_sbuf);
+    return rc;
+}
+
+int flb_chunk_trace_filter(struct flb_chunk_trace *tracer, void *pfilter, struct flb_time *tm_start, struct flb_time *tm_end, char *buf, size_t buf_size)
+{
+    msgpack_packer mp_pck;
+    msgpack_sbuffer mp_sbuf;
+    msgpack_unpacked result;
+    msgpack_object *record = NULL;
+    int rc = -1;
+    struct flb_filter_instance *filter = (struct flb_filter_instance *)pfilter;
+    flb_sds_t tag = flb_sds_create("trace");
+    struct flb_time tm;
+    size_t off = 0;
+    int records = 0;
+
+
+    if (tracer == NULL) {
+        goto tracer_error;
+    }
+
+    msgpack_sbuffer_init(&mp_sbuf);
+    msgpack_packer_init(&mp_pck, &mp_sbuf, msgpack_sbuffer_write);
+
+    msgpack_pack_array(&mp_pck, 2);
+    flb_pack_time_now(&mp_pck);
+    if (filter->alias == NULL) {
+        msgpack_pack_map(&mp_pck, 6);
+    }
+    else {
+        msgpack_pack_map(&mp_pck, 7);	
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "type", strlen("type"));
+    rc = msgpack_pack_int(&mp_pck, FLB_CHUNK_TRACE_TYPE_FILTER);
+    if (rc == -1) {
+        goto sbuffer_error;
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "start_time", strlen("start_time"));
+    //msgpack_pack_double(&mp_pck, flb_time_to_double(tm_start));
+    flb_time_append_to_msgpack(tm_start, &mp_pck, FLB_TIME_ETFMT_INT);
+    msgpack_pack_str_with_body(&mp_pck, "end_time", strlen("end_time"));
+    //msgpack_pack_double(&mp_pck, flb_time_to_double(tm_end));
+    flb_time_append_to_msgpack(tm_end, &mp_pck, FLB_TIME_ETFMT_INT);
+
+    msgpack_pack_str_with_body(&mp_pck, "trace_id", strlen("trace_id"));
+    msgpack_pack_str_with_body(&mp_pck, tracer->trace_id, strlen(tracer->trace_id));
+
+    
+    msgpack_pack_str_with_body(&mp_pck, "plugin_instance", strlen("plugin_instance"));
+    rc = msgpack_pack_str_with_body(&mp_pck, filter->name, strlen(filter->name));
+    if (rc == -1) {
+        goto sbuffer_error;
+    }
+    
+    if (filter->alias != NULL) {
+        msgpack_pack_str_with_body(&mp_pck, "plugin_alias", strlen("plugin_alias"));
+        msgpack_pack_str_with_body(&mp_pck, filter->alias, strlen(filter->alias));
+    }
+
+    msgpack_pack_str_with_body(&mp_pck, "records", strlen("records"));
+
+    msgpack_unpacked_init(&result);
+
+    if (buf_size > 0) {
+        do {
+            rc = msgpack_unpack_next(&result, buf, buf_size, &off);
+            if (rc != MSGPACK_UNPACK_SUCCESS) {
+                flb_error("unable to unpack record");
+                goto unpack_error;
+            }
+            records++;
+        } while (rc == MSGPACK_UNPACK_SUCCESS && off < buf_size);
+
+        msgpack_pack_array(&mp_pck, records);
+        off = 0;
+        do {
+            rc = msgpack_unpack_next(&result, buf, buf_size, &off);
+            if (rc != MSGPACK_UNPACK_SUCCESS) {
+                flb_error("unable to unpack record");
+                goto unpack_error;
+            }
+            flb_time_pop_from_msgpack(&tm, &result, &record);
+
+            msgpack_pack_map(&mp_pck, 2);
+            msgpack_pack_str_with_body(&mp_pck, "timestamp", strlen("timestamp"));
+            flb_time_append_to_msgpack(&tm, &mp_pck, FLB_TIME_ETFMT_INT);
+            msgpack_pack_str_with_body(&mp_pck, "record", strlen("record"));
+            msgpack_pack_object(&mp_pck, *record);
+
+        } while (rc == MSGPACK_UNPACK_SUCCESS && off < buf_size);
+    }
+
+    in_emitter_add_record(tag, flb_sds_len(tag), mp_sbuf.data, mp_sbuf.size,
+                          tracer->ctxt->input);
+    
+    rc = 0;
+
+unpack_error:
+    msgpack_unpacked_destroy(&result);
+sbuffer_error:
+    msgpack_sbuffer_destroy(&mp_sbuf);
+tracer_error:
+    flb_sds_destroy(tag);
+    return rc;
+}