How golang garbage collection works

It is a Mark-and-Sweep GC. Phases Mark Stop-the-World: Set write barrier (to know how much was allocated during maark phase) Concurrent: Mark all memory which is still in use by the app Stop-the-World: Remove write barrier Concurrent: Sweep (it actually happens on new allocations) Details/Algorithm It’s a tri-color process: grey for objects to check, black …

Linux. History of system usage

There is a way to monitor your system usage and log to a file easily — with atop command. On my Ubuntu it can be installed with

sudo apt-get install atop

1	sudo apt-get install atop

It can be started as a top command — it’s output almost the same. But the most interesting feature is that it starts a daemon that logs …

Style guide for golang project file structure

Highly recommended. It has a couple of odds like not explaining about /test dir (I assume, they wanted to say to put there only e2e and other integrational tests, not unit-test as the latter should be kept together with the code itself in files like …_test.go). But it’s awesome in general. https://github.com/golang-standards/project-layout

Golang patterns. Worker pool

func worker(id int, jobs <-chan int, results chan<- int) {
    for j := range jobs {
        fmt.Println("worker", id, "started  job", j)
        time.Sleep(time.Second)
        fmt.Println("worker", id, "finished job", j)
        results <- j * 2
    }
}

func main() {

    // In order to use our pool of workers we need to send
    // them work and collect their results. We make 2
    // channels for this.
    jobs := make(chan int, 100)
    results := make(chan int, 100)

    // This starts up 3 workers, initially blocked
    // because there are no jobs yet.
    for w := 1; w <= 3; w++ {
        go worker(w, jobs, results)
    }

    // Here we send 5 `jobs` and then `close` that
    // channel to indicate that's all the work we have.
    for j := 1; j <= 5; j++ {
        jobs <- j
    }
    close(jobs)

    for a := 1; a <= 5; a++ {
        <-results
    }
}

func worker(id int, jobs <-chan int, results chan<- int) {

for j := range jobs {

fmt.Println("worker", id, "started job", j)

time.Sleep(time.Second)

fmt.Println("worker", id, "finished job", j)

results <- j * 2

}

func main() {

// In order to use our pool of workers we need to send

// them work and collect their results. We make 2

// channels for this.

jobs := make(chan int, 100)

results := make(chan int, 100)

// This starts up 3 workers, initially blocked

// because there are no jobs yet.

for w := 1; w <= 3; w++ {

go worker(w, jobs, results)

}

// Here we send 5 `jobs` and then `close` that

// channel to indicate that's all the work we have.

for j := 1; j <= 5; j++ {

jobs <- j

}

close(jobs)

for a := 1; a <= 5; a++ {

<-results

}

Further reading: https://gobyexample.com/worker-pools

PgBouncer and prepared statements

In our system, we use connection pooler called PgBouncer as a proxy to PostgreSQL server. PgBouncer has two main modes Session pooling mode This mode is less performant (it’s a default mode). When a client connects, a server connection will be assigned to it for the whole duration it stays connected. So it does not …

Golang: testing http and grpc servers

HTTP server is quite easy to test — here is a nice video about it:

Count lines in git repository

git ls-files | xargs wc -l

1	git ls-files \| xargs wc -l

https://stackoverflow.com/questions/4822471/count-number-of-lines-in-a-git-repository

Algorithms to find fastest route through network

This is a great article on that — https://www.redblobgames.com/pathfinding/a-star/introduction.html (in russain). It tells about Dijkstra, Breadth-first, Greedy and A* algorithms. It’s all about optimization of one travel. To optimize your entire network you should try to use Aint Colony algorithm — http://rain.ifmo.ru/cat/data/theory/unsorted/ant-algo-2006/article.pdf (it’s in russian). Another link to this PDF file — aca

How go code is being compiled to assembler code

https://go.godbolt.org/z/31FyJ3 I was interested in comparing line 15 vs line 17 of the following code:

package main

import "fmt"

type kv struct {
	key   []byte
	value []byte
}

type sliceMap []kv

func (sm *sliceMap) Add(k, v []byte) {
	kvs := *sm
	if cap(kvs) > len(kvs) {
		kvs = kvs[:len(kvs)+1]
	} else {
		kvs = append(kvs, kv{})
	}
	kv := &kvs[len(kvs)-1]
	kv.key = append(kv.key[:0], k...)
	kv.value = append(kv.value[:0], v...)
	*sm = kvs
}

func main() {
	sm := sliceMap{}
	sm.Add([]byte("foo"), []byte("bar"))
	fmt.Println(sm)
}

package main

import "fmt"

type kv struct {

key []byte

value []byte

}

type sliceMap []kv

func (sm *sliceMap) Add(k, v []byte) {

kvs := *sm

if cap(kvs) > len(kvs) {

kvs = kvs[:len(kvs)+1]

} else {

kvs = append(kvs, kv{})

}

kv := &kvs[len(kvs)-1]

kv.key = append(kv.key[:0], k...)

kv.value = append(kv.value[:0], v...)

*sm = kvs

}

func main() {

sm := sliceMap{}

sm.Add([]byte("foo"), []byte("bar"))

fmt.Println(sm)

}

E2E tests in go

I tried it via ginkgo and gomega. It has Agouti with WebDriver support out of the box, but I didn’t use it as we have grpc API. That’s the example (table tests):

var _ = Describe("GetRoute", func() {
	var (
		srClient sr.Client
		srErr    error
	)

	JustBeforeEach(func() {
		srClient, srErr = GetClient()
	})

	DescribeTable("positive cases",
		func(fromGeoPointID, toGeoPointID int, expectedUID string, expectedCostGreaterThan float64) {
			Expect(srErr).NotTo(HaveOccurred())
			Expect(srClient).NotTo(BeNil())

			response, err := srClient.GetRoute(context.TODO(), &sr.GetRouteRequest{
				FromGeoPointId: int64(fromGeoPointID),
				ToGeoPointId:   int64(toGeoPointID),
				Quantity:       1,
				Volume:         1,
				Weight:         1,
			})

			Expect(err).To(Succeed())
			Expect(response).NotTo(BeNil())
			Expect(response.RouteCost).NotTo(BeNil())
			Expect(response.RouteCost.RouteUid).To(HaveSuffix(expectedUID))
			Expect(response.RouteCost.GetCost()).To(BeNumerically(">", expectedCostGreaterThan))
		},
		Entry("from(geo=3)->to(geo=2)", 3, 2, "-3-2", float64(10)),
	)

var _ = Describe("GetRoute", func() {

var (

srClient sr.Client

srErr error

)

JustBeforeEach(func() {

srClient, srErr = GetClient()

})

DescribeTable("positive cases",

func(fromGeoPointID, toGeoPointID int, expectedUID string, expectedCostGreaterThan float64) {

Expect(srErr).NotTo(HaveOccurred())

Expect(srClient).NotTo(BeNil())

response, err := srClient.GetRoute(context.TODO(), &sr.GetRouteRequest{

FromGeoPointId: int64(fromGeoPointID),

ToGeoPointId: int64(toGeoPointID),

Quantity: 1,

Volume: 1,

Weight: 1,

})

Expect(err).To(Succeed())

Expect(response).NotTo(BeNil())

Expect(response.RouteCost).NotTo(BeNil())

Expect(response.RouteCost.RouteUid).To(HaveSuffix(expectedUID))

Expect(response.RouteCost.GetCost()).To(BeNumerically(">", expectedCostGreaterThan))

Entry("from(geo=3)->to(geo=2)", 3, 2, "-3-2", float64(10)),

)