Skip to content

bad-logic/distributed-tracing

Repository files navigation

How to run application !!!

Setup Everything with scripts

# run apps
$ chmod +x run.sh
$ ./run.sh
# create tables in db
$ chmod +x migrations.sh
$ ./migrations.sh run

OR Run each infrastructure manually First run kafka

docker-compose -f deploy/kafka-compose.yaml up -d

Second run logging containers

docker-compose -f deploy/logs-compose.yaml up -d

Third run application containers

docker-compose -f deploy/docker-compose.yaml up

Fourth run migration to create tables in database

$ chmod +x migrations.sh
$ ./migrations.sh run

Run bash test.sh to send some logs/traces to signoz

Running Applications
product microservice -> localhost:8082/products
order microservice -> localhost:8083/order/
kafka ui -> localhost:8081
Signoz -> localhost:8080
consumer node
consumer cpp

Api Documentation

Screenshots

signoz

DISTRIBUTED TRACING

In distributed systems, a single operation may traverse across multiple software components. each software components may produce traces/logs that are uniquely identifiable across the participating systems. so it is very difficult to correlate those traces/logs.

In order to correlate those traces/logs, they must share a context that tells them that they are part of a specific transaction/operation.

Trace context is something that can be passed along with the request, that tells the participating system that it is a part of a specific operation.

Trace context is split into two individual propagation fields supporting interoperability and vendor-specific extensibility:

traceparent header : represents the incoming request in a tracing system in a common format, understood by all vendors.

Format: {version}-{trace_id}-{parent_id}-{trace_flags}

version => 2 HEXDIGLC (Hexadecimal digits lowercase)

The current specification assumes the version is set to 00. Version ff is forbidden

trace_id => 32 HEXDIGLC (16 bytes array identifier. All zeroes forbidden)

All traces belonging to the same transaction will have same trace_id. it is a unique identifier across the distributed system.

how to generate trace_id ?

parent_id => 16 HEXDIGLC (8 bytes array identifier. All zeroes forbidden)

It is the id of the caller. In some tracing system it may also be known as the span_id.

trace_flags => 2 HEXDIGLC (8 bit flags.)

The current version of this specification (00) only supports a single flag called sampled.

static final byte FLAG_SAMPLED = 1; // 00000001
boolean sampled = (traceFlags & FLAG_SAMPLED) == FLAG_SAMPLED;

Ex:

00-4bf92f3577b34da6a3ce929d0e0e4736-00f067aa0ba902b7-01
base16(version) = 00
base16(trace-id) = 4bf92f3577b34da6a3ce929d0e0e4736
base16(parent-id) = 00f067aa0ba902b7
base16(trace-flags) = 01  // sampled
00-4bf92f3577b34da6a3ce929d0e0e4736-00f067aa0ba902b7-00
base16(version) = 00
base16(trace-id) = 4bf92f3577b34da6a3ce929d0e0e4736
base16(parent-id) = 00f067aa0ba902b7
base16(trace-flags) = 00  // not sampled

tracestate header (optional): includes the parent in a potentially vendor-specific format

Read more...