Understanding the logic of WaitGroups - go

I'd like to understand if my logic around WaitGroups is correct and to see if there is a more efficient way of structuring my code. The aim is to perform the tasks as fast as possible.
My code populates a _urls channel which is populated via stdin. Then I'm spinning up two WaitGroups, one which reads from this _urls channel, and the other which reads from a _downloads channel, which is fed from a goroutine in the first WaitGroup.
Essentially the code looks like this:
// declare channels
_urls := make(chan string)
_downloads := make(chan string)
// first waitgroup with 2 goroutines
var wg sync.WaitGroup
for i := 0; i < concurrency; i++ {
wg.Add(2)
go func() {
defer wg.Done()
for url := range _urls {
// perform GET request and inspect the responseBody
}
}()
go func() {
defer wg.Done()
for url := range _urls {
// perform a HEAD request to look for a certain file
// if the file exists, send to the _downloads channel
_downloads <- url
}
}()
}
// second waitgroup with 1 goroutine
var dwg sync.WaitGroup
for i := 0; i < concurrency; i++ {
dwg.Add(1)
go func() {
defer wg.Done()
for url := range _downloads {
// perform the download
}
}()
}
My concern is around whether this is an efficient way to feed the _downloads channel, or would it make more sense to just perform the download within the first WaitGroup.

I did something similar to this with the worker pool pattern, https://gobyexample.com/worker-pools. Probably the right direction if you are looking to maximizing concurrency.
It abstracts jobs using go interfaces, so it could be a HEAD, GET, Download, or whatever else happens to make sense in the future. A scheduler sends jobs to a dispatcher that manages the worker pool and sends results back.
Here is a link to the README and code.
It uses wait groups to track the number of active workers, not jobs. Workers execute a for {} loop and only exit when they read true from a done channel. In this case, the use of the wait group is for a graceful shutdown. In your example, many of the workers could be doing long downloads. So your shutdown logic could wait for N jobs to be left before shutting down.
It may be overkill for your use cases.

Related

Stopping other goroutines conditionally

I am in doubt whether all of my spawned goroutines are dying after doing their assigned work.
I have to make two HTTP calls(always), but based on a flag, read the response from either one of them.
what I have done so far is ->
var result error
resultChannel := make(chan error)
var wg sync.WaitGroup
wg.Add(1) // only adding 1, as I don't need to wait for other to complete.
go func() {
_, err := // HTTP call ONE
if flagIsTrue {
defer wg.Done()
resultChannel <- err
}
}()
go func() {
_, err := // HTTP call TWO
if !flagIsTrue {
defer wg.Done()
resultChannel <- err
}
}()
go func() {
wg.Wait()
close(resultChannel)
}()
for err := range resultChannel {
result = err
}
Hence, I will wait for the corresponding call, and listen to its response only. This is working well, but since the app is deployed on the server, where I guess the main goroutine won't die(henceforth killing other goroutines), my main concern is whether the other ignorable thread will die or not after it will get the response from HTTP call(afaik, we need to tell go that a goroutine needs to die).
My concerns:
The assumption(true acc to me) that the main thread does not terminate after serving one of these calls.
Will the ignorable(response is, but necessary to trigger the API call) thread die or not?
Should I use a select case to handle this, if yes then how(other suggestions are welcome)?
If the flagIsTrue is set before creating the goroutines, then only one of the goroutines will be able to write to the channel. The other one will not attempt to write to the channel, and thus will terminate.
You could simply move the check for the flag outside, and create one goroutine based on the flag.

Goroutine example - Donovan/Kernighan book - explanation needed

The following example is taken from the Donovan/Kernighan book:
func makeThumbnails6(filenames <-chan string) int64 {
sizes := make(chan int64)
var wg sync.WaitGroup // number of working goroutines
for f := range filenames {
wg.Add(1)
// worker
go func(f string) {
defer wg.Done()
thumb, err := thumbnail.ImageFile(f)
if err != nil {
log.Println(err)
return
}
info, _ := os.Stat(thumb) // OK to ignore error
sizes <- info.Size()
}(f)
}
// closer
go func() {
wg.Wait()
close(sizes)
}()
var total int64
for size := range sizes {
total += size
}
return total
}
And the book states:
"These two operations, wait and close, must be concurrent with the
loop over sizes. Consider the alternatives: if the wait operation were
placed in the main goroutine before the loop, it would never end"
This is what I do not understand - if they are not put in separate goroutine, then wg.Wait will block the main goroutine, so close(sizes) will happen when all other goroutines finished. Closing the sizes channel will still allow the loop to read all the already sent messages/ from channel, right?
Closing the sizes channel will still allow the loop to read all the already sent messages/ from channel, right?
Yes, but that's not the problem. All goroutines are waiting to read from channels (and therefore, nobody will ever write to them). So the process will deadlock if sizes is unbuffered. For the workers to complete, something needs to read from it. For wg.Wait() to complete, the workers need to complete.
But also, the range sizes can't complete (eg find an empty, closed channel) until close(sizes) happens, which can't complete until the workers are complete (because they're the ones writing to sizes).
So it's wg.Wait() and close(sizes) both have to complete before range sizes that happen concurrently.
Unbuffered channnel sizes would block all other routines to write into it. wg.Wait() would never end because all routines are blocked in sizes <-info.Size().
So it's nessary that wait and close are concurrent with the loop over sizes.

A case of `all goroutines are asleep - deadlock!` I can't figure out why

TL;DR: A typical case of all goroutines are asleep, deadlock! but can't figure it out
I'm parsing the Wiktionary XML dump to build a DB of words. I defer the parsing of each article's text to a goroutine hoping that it will speed up the process.
It's 7GB and is processed in under 2 minutes in my machine when doing it serially, but if I can take advantage of all cores, why not.
I'm new to threading in general, I'm getting a all goroutines are asleep, deadlock! error.
What's wrong here?
This may not be performant at all, as it uses an unbuffered channel, so all goroutines effectively end up executing serially, but my idea is to learn and understand threading and to benchmark how long it takes with different alternatives:
unbuffered channel
different sized buffered channel
only calling as many goroutines at a time as there are runtime.NumCPU()
The summary of my code in pseudocode:
while tag := xml.getNextTag() {
wg.Add(1)
go parseTagText(chan, wg, tag.text)
// consume a channel message if available
select {
case msg := <-chan:
// do something with msg
default:
}
}
// reading tags finished, wait for running goroutines, consume what's left on the channel
for msg := range chan {
// do something with msg
}
// Sometimes this point is never reached, I get a deadlock
wg.Wait()
----
func parseTagText(chan, wg, tag.text) {
defer wg.Done()
// parse tag.text
chan <- whatever // just inform that the text has been parsed
}
Complete code:
https://play.golang.org/p/0t2EqptJBXE
In your complete example on the Go Playground, you:
Create a channel (line 39, results := make(chan langs)) and a wait-group (line 40, var wait sync.WaitGroup). So far so good.
Loop: in the loop, sometimes spin off a task:
if ...various conditions... {
wait.Add(1)
go parseTerm(results, &wait, text)
}
In the loop, sometimes do a non-blocking read from the channel (as shown in your question). No problem here either. But...
At the end of the loop, use:
for res := range results {
...
}
without ever calling close(results) in exactly one place, after all writers finish. This loop uses a blocking read from the channel. As long as some writer goroutine is still running, the blocking read can block without having the whole system stop, but when the last writer finishes writing and exits, there are no remaining writer goroutines. Any other remaining goroutines might rescue you, but there are none.
Since you use the var wait correctly (adding 1 in the right place, and calling Done() in the right place in the writer), the solution is to add one more goroutine, which will be the one to rescue you:
go func() {
wait.Wait()
close(results)
}()
You should spin off this rescuer goroutine just before entering the for res := range results loop. (If you spin it off any earlier, it might see the wait variable count down to zero too soon, just before it gets counted up again by spinning off another parseTerm.)
This anonymous function will block in the wait variable's Wait() function until the last writer goroutine has called the final wait.Done(), which will unblock this goroutine. Then this goroutine will call close(results), which will arrange for the for loop in your main goroutine to finish, unblocking that goroutine. When this goroutine (the rescuer) returns and thus terminates, there are no more rescuers, but we no longer need any.
(This main code then calls wait.Wait() unnecessarily: Since the for didn't terminate until the wait.Wait() in the new goroutine already unblocked, we know that this next wait.Wait() will return immediately. So we can drop this second call, although leaving it in is harmless.)
The problem is that nothing is closing the results channel, yet the range loop only exits when it closes. I've simplified your code to illustrate this and propsed a solution - basically consume the data in a goroutine:
// This is our producer
func foo(i int, ch chan int, wg *sync.WaitGroup) {
defer wg.Done()
ch <- i
fmt.Println(i, "done")
}
// This is our consumer - it uses a different WG to signal it's done
func consumeData(ch chan int, wg *sync.WaitGroup) {
defer wg.Done()
for x := range ch {
fmt.Println(x)
}
fmt.Println("ALL DONE")
}
func main() {
ch := make(chan int)
wg := sync.WaitGroup{}
// create the producers
for i := 0; i < 10; i++ {
wg.Add(1)
go foo(i, ch, &wg)
}
// create the consumer on a different goroutine, and sync using another WG
consumeWg := sync.WaitGroup{}
consumeWg.Add(1)
go consumeData(ch,&consumeWg)
wg.Wait() // <<<< means that the producers are done
close(ch) // << Signal the consumer to exit
consumeWg.Wait() // << Wait for the consumer to exit
}

Golang Multiple Channel Write/Receive Ordering

My specific issue is that I have an unbuffered channel and am spawning multiple goroutines bounded with a semaphore to perform work:
func main() {
sem := make(chan struct{}, 10) // allow ten concurrent parsers
wg := &sync.WaitGroup{}
wg.Add(1)
DoSomething("http://example.com", sem, wg)
wg.Wait()
// all done
}
func DoSomething(u string, sem chan struct{}, wg *sync.WaitGroup) {
defer wg.Done()
sem <- struct{}{} // grab
defer func() { <-sem }() // release
var newSomethings []string
// ...
for u := range newSomethings {
wg.Add(1)
go DoSomething(u)
}
}
If there are multiple DoSomething goroutines on the stack, blocked on the sem write (or inversely on a read) When a write happens is there any ordering to which go routine gets through with the write?? I would guess it were random but I could imagine:
it is random
writes/receives happen in the order they are registered
implementation dependent
I looked at a couple of resources and was unable to find a solution:
https://github.com/golang/go/issues/247
https://golang.org/ref/spec#Receive_operator
https://golang.org/ref/spec#Channel_types
I'm wondering if this is undefined and/or implementation dependent, or if this logic is located and defined somewhere within go core?
The order that goroutines blocked on a send operation are serviced is not defined, but it's implemented as a FIFO. You can see the implementation in runtime/chan.go, which uses a linked list to track the channel's senders and receivers.
We can try to make an example showing the effective ordering like so:
func main() {
ch := make(chan int)
ready := make(chan int)
for i := 0; i < 10; i++ {
i := i
go func() {
ready <- 1
ch <- i
}()
<-ready
runtime.Gosched()
}
for i := 0; i < 10; i++ {
v := <-ch
if i != v {
panic("out of order!")
}
fmt.Println(v)
}
}
https://play.golang.org/p/u0ukR-5Ptw4
This still isn't technically correct, because there's no way to observe blocking on a send operation, so there's still a race between the ready send and the send to ch on the next line. We can try to eliminate that with the runtime.Gosched call here, or even a time.Sleep, but without explicit synchronization there's no guarantee of a "happens before" relationship.
Regardless, this queues up the goroutines and shows the expected output order, and if they weren't queued up already, it would be more likely to process the values out of order.
You can see by this example that we can't truly determine the order that the goroutines are queued up, it is almost always non-deterministic, and therefore reasoning about this isn't usually useful in practice.

Why is my Golang Channel Write Blocking Forever?

I've been attempting to take a swing at concurrency in Golang by refactoring one of my command-line utilities over the past few days, but I'm stuck.
Here's the original code (master branch).
Here's the branch with concurrency (x_concurrent branch).
When I execute the concurrent code with go run jira_open_comment_emailer.go, the defer wg.Done() never executes if the JIRA issue is added to the channel here, which causes my wg.Wait() to hang forever.
The idea is that I have a large amount of JIRA issues, and I want to spin off a goroutine for each one to see if it has a comment I need to respond to. If it does, I want to add it to some structure (I chose a channel after some research) that I can read from like a queue later to build up an email reminder.
Here's the relevant section of the code:
// Given an issue, determine if it has an open comment
// Returns true if there is an open comment on the issue, otherwise false
func getAndProcessComments(issue Issue, channel chan<- Issue, wg *sync.WaitGroup) {
// Decrement the wait counter when the function returns
defer wg.Done()
needsReply := false
// Loop over the comments in the issue
for _, comment := range issue.Fields.Comment.Comments {
commentMatched, err := regexp.MatchString("~"+config.JIRAUsername, comment.Body)
checkError("Failed to regex match against comment body", err)
if commentMatched {
needsReply = true
}
if comment.Author.Name == config.JIRAUsername {
needsReply = false
}
}
// Only add the issue to the channel if it needs a reply
if needsReply == true {
// This never allows the defered wg.Done() to execute?
channel <- issue
}
}
func main() {
start := time.Now()
// This retrieves all issues in a search from JIRA
allIssues := getFullIssueList()
// Initialize a wait group
var wg sync.WaitGroup
// Set the number of waits to the number of issues to process
wg.Add(len(allIssues))
// Create a channel to store issues that need a reply
channel := make(chan Issue)
for _, issue := range allIssues {
go getAndProcessComments(issue, channel, &wg)
}
// Block until all of my goroutines have processed their issues.
wg.Wait()
// Only send an email if the channel has one or more issues
if len(channel) > 0 {
sendEmail(channel)
}
fmt.Printf("Script ran in %s", time.Since(start))
}
The goroutines block on sending to the unbuffered channel.
A minimal change unblocks the goroutines is to create a buffered channel with capacity for all issues:
channel := make(chan Issue, len(allIssues))
and close the channel after the call to wg.Wait().

Resources